N Innovations in Clinical Trial Design

IN THIS ISSUE: FOCUS ON CUTTING- EDGE RESEARCH
SUMMER 2008
Innovations
in Clinical
Trial Design
Adaptive Designs That Are
Both Flexible and Rigorous
BY STUART POCOCK
N
ERI has successfully carried
out many major randomized
clinical trials across a wide
range of diseases and types
of intervention. This wealth of experience
means NERI excels in the rigorous planning
of trials based on protocols developed at
the inception of a project.
One major issue in the development of any
trial protocol is how many patients must be
followed over what mean period of time
for the results to have statistical validity.
For some trials a track record of previous
research exists which provides the information needed for performing sensible
statistical power calculations that yield the
answer: the trial needs X patients.
Other randomized trials are more radically
innovative so that (dare one say it) quite
a lot of guesswork goes into such power
calculations. Lack of prior research on the
disease, on possible interventions and/or
on outcome measures means you simply
don’t know what will happen to the control
group nor what magnitude of treatment
difference is achievable and clinically relevant. This is just the circumstance where
Adaptive Designs can come to the rescue.
They are gaining in popularity, but still
need rigorous planning. NERI has put
together a plenary session on Adaptive
Designs for the DIA conference held in
Boston in June 2008. continued on page 19
STUART POCOCK has been a visiting chief scientist at NERI since 1996. He is professor of Medical Statistics at the London School of
Hygiene and Tropical Medicine, and is an internationally recognized biostatistical expert in the practice and methodology of randomized
controlled trials. On each of his frequent forays to the US for trial collaborations, conferences, or lecturing, he visits NERI to consult
on the whole range of NERI public health research, both trials and observational studies. Stuart has a current research focus on adaptive
designs, both in practical experiences and methodological insights. “My collaborations at NERI are always enjoyable because we are
delivering high-quality studies of great public health importance,” Stuart says. “We are covering a wide range of major health issues, maintaining rigorous
standards and stimulating useful improvements in methodology. NERI has a ‘can-do’ philosophy, especially in nurturing collaborations and networks.”
McKinlay is Honored
New England Research Institutes
9 Galen Street
Watertown, MA 02472
(617) 923-7747
FAX (617) 926-8246
www.neriscience.com
John McKinlay, NERI’s Senior Vice-President and Scientific Director,
is to be honored in multiple ways this year for his many contributions to the fields of epidemiology, sociology, and public health.
This June he will deliver the Matilda White Riley Lecture at the
National Institutes of Health. His topic: “Upstream Determinants
of Downstream Health Disparities.” This lecture, which honors
No research without
therapeutic or policy benefit
the noted NIH sociologist, is awarded annually in recognition of
outstanding research.
• Institute for Clinical Trials
and Registries
• Institute for Studies on Aging
• Institute for Health Policy
and Services Research
In August at the annual meeting of the American Sociological
Association in Boston, he will be honored by colleagues with a
prestigious award for “Lifetime Distinguished Contributions to the Practice of Sociology.”
He has also been invited to participate at the 40th anniversary meeting of the Medical Sociology
• Institute for Community
Health Studies
Group in Brighton, England. McKinlay was instrumental in founding this important professional
• Institute for New Media and
Communications Research
Although such honors are often bestowed on scientists after they have ended their full-time research,
group more than 40 years ago, while he was a student at Aberdeen University in Scotland.
this is not the case here. McKinlay continues to work on multiple research projects simultaneously—
• Center for Scientific Integrity
as he had done throughout his career—and continues to produce an impressive output of publications.
• Center for Data Systems Technology
He remains a scientific leader in the fields of epidemiology, endocrinology, urology, clinical decision
• Center for Statistical Analysis
and Research
making and health services research.
• Center for Qualitative Research
• Survey Research Center
New Leader of Clinical Research
• Center for Health Economics
and Outcomes Research
NERI’s work as a Contract Research Organization servicing pharmaceutical, biotechnology, and medical device companies has grown
Editor: Stephen Braun
Designer: Carol Dirga
rapidly. In order to better focus and direct this key business unit,
NERI in February announced the promotion of Rebecca Li to Vice
This issue’s contributors:
Andre Araujo
Karen Barzilay
Joanne Gormley
Rebecca Li
Carol Link
Lisa Marceau
Sarah McGraw
John McKinlay
Sonja McKinlay
Patti Nash
Stuart Pocock
Ray Rosen
Sandi Siami
Lynn Sleeper
Sharon Tennstedt
To order additional copies, or
obtain further information on
NERI projects and scientists, please
contact [email protected]
President of Clinical Research. Li and her team will work to expand
NERI’s role as a private-sector CRO, while building on NERI’s 20-year
reputation as a leading public-sector research organization.
Reporting directly to NERI President, Sonja McKinlay, Li will manage
clinical research operations and all aspects of the company’s clinical
trials, registries, and related initiatives. “We are thrilled that Dr. Li will have a direct impact on our
next decade of growth,” McKinlay says. “As a scientifically-driven organization, we conduct clinical
research with integrity and passion.”
Prior to her new VP role, Li served as Principal Research Scientist and Clinical Director at NERI.
She continues to lead, as Principal Investigator, a global Phase III clinical trial on a potential treatment
for heart failure. The trial involves thousands of patients and more than 200 clinical centers worldwide.
Before joining NERI, Li spent 14 years in the private sector, most recently at the Director level at
Wyeth Research working in the Translational Research Department. In that position, Li coordinated
efforts to incorporate pharmacogenomics into clinical trials. Prior to that work at Wyeth, Li led a
preclinical lab group in the Discovery Department focused on orthopedic tissue engineering.
Li is also an Adjunct Associate Professor at Brown University in the Department of Pharmacology,
Physiology, and Biotechnology. Li has a doctorate from Johns Hopkins University in Biomolecular
and Chemical Engineering.
NEW RESEARCH
Dental Disease in Children
A New Study Finds Much
Room For Improvement
Quick: What’s the most common chronic
childhood disease? Asthma? Allergies?
Attention Deficit Disorder?
Answer: None of the above.
T
ILLUSTRATION: DAN FREY
he true culprit is dental caries—
the demineralization and erosion
of teeth caused by bacteria. This
disease is five times more common than
asthma. If not treated, caries results in
cavities, pain, and infection that can
have serious consequences for a child’s
overall health. NERI recently completed
a large survey of dental health—including the incidence of caries—in young
children in order to get a clearer picture
of the extent of the problem.
NERI researchers collaborated with
the Boston University School of
Dental Medicine to complete the study,
commissioned by the Catalyst Institute
of Delta Dental of Massachusetts. The
goal: to assess the oral health status of
children in kindergarten, third, and
sixth grades in Massachusetts. Data were
collected via open-mouth examinations
of almost 6000 children by BU dentists
and dental hygienists using a standard
protocol for school age children. These
data were supplemented by parent
report by questionnaire. Using complex
survey sampling methods, schools and
classrooms were randomly selected to
provide reliable estimates for the state
as a whole as well as for each of the
state’s 14 counties. Results of the study
were released in a well-attended and
well-received presentation at the
Massachusetts State House in January
2008.
A significant portion of Massachusetts
children suffer from dental caries, and
many children start school with this
Between 4% and 5% of school
children are in pain from caries,
which may impede their learning.
disease. One-quarter of kindergarten
children, 40% of third-graders and
one-third of sixth-graders were found
to have caries. The study found that
between 4% and 5% of children in
continued on page 4
3
NEW RESEARCH
Dental Disease is a Serious Health Issue for Massachusetts Children
Percent and number of children
50%
41%
40%
30%
29,114
34%
History of cavities
Untreated cavities
24,575
28%
19,130
20%
17%
15%
12,375
10%
9,942
11%
7,949
0%
Kindergarten
3rd Grade
school are in pain from caries, which
may impede their learning. Although
the rates have declined in third graders
since the state’s last oral health survey
in 2003, substantial disease remains.
6th Grade
sixth graders had received dental
sealants, with wide variability across
the 14 counties.
Further, significant racial, ethnic, and
socioeconomic disparities exist within
all oral health indicators, at each grade
level, and across the state’s 14 counties.
In general, the rates of dental caries,
untreated decay, and pain were higher in
children from low-income families and
families without a regular dentist as well
as in racial and ethnic minority children.
The results of this study show that
more needs to be done in Massachusetts
to prevent oral disease, reduce untreated
disease, and eliminate oral health disparities. The Catalyst Institute will
collaborate with other stakeholders to use
these study results to provide a framework for improviing dental care and
access to care for the state’s children and
to monitor statewide progress toward
Healthy People 2020 oral health goals.
Dental caries and resulting cavities can
be prevented and easily treated with
access to timely dental care. The results
of this study indicate that although some
progress has been made in the past 4
years, children with the most oral disease are the least likely to receive dental
sealants, a safe and effective preventive
measure. Only about half of third and
For more information about the
Catalyst Institute and this study:
www.catalystinstitute.org
NERI investigators: Sharon Tennstedt
and Anne Stoddard.
BU investigators: Michelle Henshaw
and Corinna Culler.
Catalyst Project Directors: B. Alex White
and Michael Monopoli.
4
Is The Doctor-Patient
Relationship Obsolete?
S
ociologist Talcott Parsons provided
a theoretical perspective on the
doctor-patient relationship (D-P
relationship), reflecting the situation in
the U.S. around the middle of the 20th
century, and his view influenced thinking
for the remainder of the century. The
Parsonian view is depicted in Figure 1
(on next page) and several features
should be highlighted:
• The spotlight of analysis was generally
on the doctor, who enjoyed high social
status and a dominant role as the
repository of medical knowledge and
expertise;
• Only two actors were involved in the
interaction, with the doctor acting
professionally and altruistically to
serve only the patient’s interests;
• The patient occupied a subordinate
and reciprocal role and was expected
to trust the doctor’s judgment and
follow (his) clinical recommendations.
Credat emptor (“let the buyer trust”)
was the prevailing ethos.
Some researchers and decision theorists
still employ this idealistic perspective,
even though the D-P relationship today
(within which clinical decisions occur)
bears little resemblance to earlier formulations. The world of health care in the
U.S. has shifted beneath decision theorists’ feet, producing theories with little
policy relevance and educational efforts
that are unlikely to produce desired
NEW RESEARCH
FIGURE 1
CARTOON: MARK TONRA
changes in clinical practice. Some indication of the magnitude of the transformation of US health care is evident in
the words used to describe the once
special D-P relationship: the doctor has
become “a provider,” the patient is now
a “client,” and the relationship is now
considered “an encounter.”
Some of the major new influences affecting clinical decision making within this
new client-provider encounter (C-P
Encounter) are illustrated in Figure 2
and include:
FIGURE 2
• Physicians are increasingly forced to
specialize (generalists are in short
supply) and most are full-time salaried
employees in large and increasingly
concentrated organizations;
• Corporate doctors are required to go
along with clinical guidelines and
pay-for-performance schemes if they
want promotions and salary increases;
• The spotlight is now on a knowledgeempowered patient/client who occupies
center stage and is the ultimate object
of all revenue in the U.S. profit-driven
health care system;
• Insurance companies increasingly
dictate what any clinician can actually
decide for any given case (test ordering,
referrals, prescriptions and follow-up);
• Pharmaceutical companies advertise
directly to consumers and suggest
that they should ask their providers
for specific medications. “Doctor
knows best” is no longer the prevailing
viewpoint;
continued on page 6
5
NEW RESEARCH
• Widely publicized reports of financial
kickbacks and clinical malpractice,
and recognition that doctors may
now serve several masters, appear to
have eroded trust in the profession of
medicine (as with car repairs, caveat
emptor is the emerging ethos). Popular
media no longer portrays doctors as
cultural heroes—compare “Marcus
Welby MD” with today’s “House” or
“Green Wing.”
U
Unlike the one-on-one, closed system
relationship of the past, clinical decision
making is now required to occur on an
ever more crowded stage. The spotlight is
now on the patient as an object of revenue
(client centered care for an objectified
condition). The patient is digitally
empowered and activated by private
(mainly pharmaceutical) interests.
In the Boston Area Community Health
(BACH) study, a random sample of community-dwelling residents, we obtained
information on urologic symptoms as
well as information about medication
use (both prescription and over the
Treatments Available, But
Patients Don’t Use Them
rologic symptoms such as
incontinence or bladder pain
are very common. A number
of medications have been proven to
help alleviate such symptoms, but data
from a large NERI study show that
these medications are not being used
by most of the people suffering from
these symptoms.
People who take medications for these
conditions may have no symptoms, mild
symptoms, or they may still have some
symptoms (though presumably less
severe than they would be without the
treatment).
Our data show that for these conditions,
about half of those on a medication are
symptomatic, while the other half have
mild or no symptoms. The prevalence of
pharmacologic treatment by condition,
gender, and level of symptoms is given
100%
MEDS
NO MEDS
80%
PERCENT
60%
40%
20%
LUTS/BPH
OAB
UI
Moderate/Severe
Symptoms
Mild Symptoms
No Symptoms
Moderate/Severe
Symptoms
Mild Symptoms
No Symptoms
Moderate/Severe
Symptoms
Mild Symptoms
No Symptoms
Moderate/Severe
Symptoms
Mild Symptoms
0%
No Symptoms
While supposedly making decisions
solely in the interest of the patient,
doctors are now required to also serve
their corporate employers (and there is
no guaranteed coincidence of interest).
The possible range of clinical actions
and costs are dictated by a patient’s
health insurance (assuming that they
have such); the government is now
essentially an onlooker, unwilling to
protect the prerogatives of doctors
and concerned to reduce the burden
of ever increasing health care costs.
Surveys reveal high levels of physician
dissatisfaction with their workplace
and complaints about administrative
encroachments on clinical autonomy.
counter). We looked at symptoms suggestive of four conditions: lower urinary
tract symptoms due to benign prostatic
hyperplasia (LUTS/BPH), urinary incontinence (UI), overactive bladder (OAB),
and painful bladder syndrome (PBS).
PBS
Figure 1: Prevalence and Use of Pharmacologic Treatments for Urologic Symptoms: MEN
6
NEW
NEWRESEARCH
AWARDS
in Figures 1 and 2. The prevalence
of each set of symptoms and level of
symptoms is the total height of each bar,
while the proportion taking (effective)
medications for those symptoms is the
portion in dark red at the top of the bar.
Less than ten percent of those with
current symptoms are taking medications
known to be at least somewhat effective
for alleviating their symptoms. Possible
reasons for not taking the medications
include: 1) for LUTS/BPH, the symptoms
may not be due to BPH so the medication would not be appropriate; and 2)
for UI, the medications available at the
time of the BACH survey were addressing urge incontinence (leaking urine
when you needed to get to the toilet
but couldn’t get there fast enough)
rather than stress incontinence (leaking
urine with some physical activity such
as coughing or sneezing).
For most of the men, the type of incontinence (urge, stress, mixed (both)) could
not be assessed. It should also be noted
that the most commonly used medication for painful bladder syndrome is an
anti-depressant and thus is not specific
to PBS.
In summary, it appears that many people
with urologic symptoms are not taking
medication known to be effective (to
some extent) for their symptoms. This
unmet need could be addressed by
increased levels of consumer education.
This research is supported by NIDDK
grant #: BACH DK56842.
100%
MEDS
NO MEDS
PERCENT
80%
60%
40%
20%
OAB
UI
Moderate/Severe
Symptoms
Mild Symptoms
No Symptoms
Moderate/Severe
Symptoms
Mild Symptoms
No Symptoms
Moderate/Severe
Symptoms
Mild Symptoms
No Symptoms
0%
PBS
“Lumping” v. “Splitting”
Urological Symptoms
T
here’s a widely accepted dogma in
urology that although clinically
distinct syndromes have been
identified, they often appear to be the
same thing. Urologic conditions like
prostatitis, benign prostatic hyperplasia/
lower urinary tract symptoms (BPH/LUTS),
urinary incontinence, and interstitial
cystitis/painful bladder syndrome
(IC/PBS) are commonly thought to be
distinct clinical entities. Continuing to
study each one in isolation, without
regard for the possibility that each
one may contain aspects of the others,
has tended to perpetuate the dogma of
separate syndromes.
This kind of diagnostic “splitting” has
been the modus operandi in everyday
clinical practice. If a patient presents
with continuing pain when his or her
bladder fills, which is relieved upon
emptying, they may be diagnosed with
prostatitis (if a young male), BPH (if an
older male), or with IC/PBS (if a female).
And some physicians may even diagnose
a urinary tract infection. With each of
these diagnoses, different treatments
may be recommended. An alternative
approach would be “lumping,” in which
it is acknowledged that symptoms often
overlap and that a given patient may
be suffering from more than a single
urologic ailment simultaneously.
continued on page 8
Figure 2: Prevalence and Use of Pharmacologic Treatments for Urologic Symptoms: WOMEN
7
NEW RESEARCH
WHAT IS BACH?
The Boston Area Community Health
(BACH) Survey is investigating a broad
range of symptoms suggestive of eight
different urologic conditions in a large
(n=5503) random sample of communitydwelling adults.
Attempting to move beyond the prevailing urologic dogma, NERI researchers
have been pursuing two distinct lines of
research. Working closely with Michael
Barry (Massachusetts General Hospital)
and using data from the Boston Area
Community Health (BACH) Survey the
researchers have found that overlapping
patterns of lower urinary tract symptoms
and pelvic pain are actually quite common.
“These findings have profound implications for clinical trials and other studies of
supposedly distinct urologic conditions,”
says John McKinlay, BACH Principal
Investigator. “How do we know what
we are actually studying? While we are
focusing on patients with x, they may
perhaps equally have y and z. Studies
of a particular condition obviously
must eliminate the overlapping effects
of other related conditions.”
urologic syndromes,” Rosen says. “The
BACH data are permitted to speak for
themselves.” Interestingly, distinct
symptom clusters were obtained for
men and women, suggesting different
gender-specific pathophysiologic pathways to urologic disease.
The BACH team is currently building
on this work and investigating the
hypothesis that urologic syndromes
once thought to be clinically distinct
may actually have a common underlying
inflammatory basis. C reactive protein
data are available for most of the BACH
sample and the results of these exciting
analyses (performed on both men and
women) may indicate that “lumping”
rather than diagnostic “splitting” should
be the preferred modus operandi in
clinical urology.
For further reading:
Barry MJ, et al. Overlap of different
urological symptom complexes in a
racially and ethnically diverse, community-based population of men and
women. BJU Int. 2008;101(1):45-51.
Cinar A, et al. Cluster Analysis and
Lower Urinary Tract Symptoms (LUTS)
in Men: Findings from the Boston Area
Community Health (BACH) Survey,
British Journal of Urology. In press.
Complementary analyses of the BACH
data, led by NERI’s Chief Scientist Ray
Rosen, have employed clustering techniques to empirically derive complexes
of symptoms which fall into statistically
distinct groups.
Hall SA, et al. Do urologic symptoms
cluster among women? Results from
the Boston Area Community Health
(BACH) Survey, British Journal of
Urology. In press.
“This work is unique in that it makes
no prior assumptions about distinct
This research is supported by NIDDK
grant #: BACH DK56842.
8
Methods in
Qualitative Research:
How many people do I need
in my qualitative study?
R
esearchers increasingly value
qualitative methods as a legitimate
form of scientific inquiry into
human behavior. With this interest
comes the benefit of increased attention
to the rigor with which these methods
are applied. Guest and his co-authors1
note, for example, that theoretical saturation, or data saturation, is commonly
referenced as a strategy for determining
study sample size; however, the specific
approach to data saturation is infrequently documented.
NERI has been applying the methodological example described by Guest
et al for determining data saturation
in its Center for Qualitative Research.
Theoretical saturation is the point at
which additional data fail to add new
concepts or themes to the existing data.
This signals the point at which the
investigators can stop data collection.2,3
Quantitative researchers typically determine sample sizes using statistical
probability theory to project the number
of observations necessary to yield estimates that are as accurate as possible.
In contrast, samples in qualitative
research typically are nonprobabilistic
and purposive. The intention in qualitative research is to select a sample that
will yield rich information or insights
NEW RESEARCH
TABLE 1. THEMES IDENTIFIED INTERVIEW TRANSCRIPT BATCH
No. of
sub-themes
added at Batch 1
(n=5)
No. of
new sub-themes
added at Batch 2
(n=5)
No.of
new sub-themes
added at Batch 3
(n= 5)
Physical signs and symptoms
13
5
2
Emotional/behavioral concerns
9
5
2
Cognitive function
3
0
0
Sexual function
6
3
5
Sleep function
5
2
2
Relationships
4
2
0
Domains of Interest
about a topic through an intensive
examination of each study case.
In this approach, researchers select the
study sample to include individuals
with characteristics that suggest a specific
awareness of, or experience with, the
topic. The sample is comprised of
individuals who are willing to talk about
their experiences. In these studies, the
respondents are the experts; they describe,
in their own words, their experiences and
their thoughts about their experiences.
The goal is to extrapolate as much of
this information as possible to identify
the full range of concepts or themes
elevant to the topic of interest.
To assess saturation, researchers review
the data (most commonly by reading
transcripts of interviews or focus groups),
listing the concepts or themes they identify. They continue to generate a list of
concepts that emerge with each successive reading, adding new themes to the
NERI developed a method
for assessing sample size that
is replicable and broadly
appropriate in the field of
qualitative research.
list. After each reading, the researchers
compare the lists, counting the number
of new concepts identified. Data saturation is determined when the number of
new themes identified falls close to zero.
Adapting the method described by
Guest et al, researchers within NERI’s
Center for Qualitative Research examined data saturation in a study of sexual
dysfunction in men. In the study, men
were asked questions about their physical
health, emotional concerns, sexual
functioning, sleep, and social relations.
Three coders reviewed an initial batch
of five interview transcripts (Batch 1)
listing the number of sub-themes that
emerged. The coders then reviewed a
second and third batch (comprised
of five transcripts each), noting new
themes that were not already on the list
with each successive batch. As shown
in Table 1, after a review of the Batch 3
transcripts, new themes continued to
arise within some domains (particularly
around the topic of sexual functioning).
As a result, we concluded that data
saturation had not been reached and that
at least one more batch of interviews
should be conducted and analyzed in
order to exhaust all possible themes.
This is an example of how NERI
researchers are continuing to advance
their relative fields of expertise. In this
case, an adaptation of a method for
assessing the “sample size” for a qualitative study yielded a method that is
replicable and broadly appropriate in
the field of qualitative research.
References
1. Guest, G, Bunce A, Johnson L.
“How Many Interviews Are Enough?
An Experiment with Data Saturation
and Variability.” Field Methods. 2006;
18(1): 59-82.
2. Morse J. “The significance of saturation.” Qualitative Health Research. 1995;
5: 147-149.
3. Patton M Q. Qualitative Research and
Evaluation Methods. 2002. Thousand
Oaks, CA., Sage Publication.
9
NERI Leads in
Clinical Registries
R
egistries were developed in the
mid-20th Century to keep track
of people with specific medical
conditions—typically tuberculosis or
cancer. The use of registries evolved into
deliberate efforts to carefully monitor
disease states and then, in more recent
years, to track the outcome of various
10
medical treatments. Registries are now
among the most important sources of
information supporting medical progress
—and NERI is a leader in the development, operation, and analysis of national
and international registries.
In recent years, registries have been used
for a variety of reasons, such as regulatory
mandates, reimbursement requirements,
and quality improvement (i.e. benchmarking or outcomes associated with
standard of care). Patient registries are
also viewed as an alternative to randomized trials in evaluating outcomes of
surgery or other medical treatments.
And registries continue to be vital in
answering important scientific questions,
such as the natural history and progression of disease.
NERI scientists and statisticians have
extensive experience designing and
implementing prospective registries,
ILLUSTRATION: DAN FREY
NEW RESEARCH
NEW RESEARCH
collaborating with a diverse group of
stakeholders ranging from government
agencies (e.g. FDA, CMS, and NIH),
industry (e.g. pharmaceutical, device,
and biotech), and the clinical community
(e.g. professional societies, academic
health centers, and clinicians). The
high-quality clinical registries developed
by NERI for its clients have proven to
be valuable resources for understanding
natural disease progression, assessing
drug/device performance, long-term
treatment evaluation, quality of life
changes, and healthcare use.
One criticism of registries is they sometimes lack monitoring or source document
verification. This is usually due to the
lack of available resources. NERI, however, has developed new ways of ensuring
data consistency, reliability, and accuracy
through remote monitoring and costeffective auditing strategies. With these
kinds of innovative techniques, NERI
believes that registry data serve very
important functions:
• Registries are able to rapidly enroll a
large number of “real-world” patients
in order to better understand natural
disease progression;
• Registries provide a cross-sectional
perspective on demographic and
clinical variables;
• Registries give investigators insight on
designing future clinical trials and a
potential network of clinical investigators;
• Registries allow for benchmarking:
comparing performance of individual
sites to the national average (or, for
international registries, comparison
of one country’s outcomes with those
of other countries in disease-specific
management);
• Registries provide estimates of outcomes such as mortality, morbidity,
quality of life, and health resource
utilization for a particular disease or
condition;
• Registries help “standardize” diagnostic
criteria by drawing a clinician’s attention to specific aspects of a particular
illness.
Standardization is essential in order for
the data collected in registries to be
comparable with other datasets or databases. Examples of standardization
include outcome or endpoint definitions,
standardized sampling or evaluation
techniques, standardized training for
participating sites, and active monitoring
of data by a central coordinating center.
NERI has developed a variety of technological tools for rapid and effective
registry implementation. For example,
the Advanced Data Entry and Protocol
Tracking (ADEPT) web-based, electronic
data capture system is customizable for
both clinical trials and registries. The
Site Management component allows for
tracking of regulatory documents and
alerts for expiring documents. The Data
Management component manages the
Analysis of registry data can be
challenging, but NERI statisticians
employ a wide range of advanced
statistical methods for interpreting
data from registry studies.
case report forms and database with
built-in validation procedures and realtime queries. The Visit Control module
allows sites to track patients and
required procedures at each visit. The
Adverse Event Manager organizes AEs
and SAEs for prompt review, reporting,
and management. The Report module
can provide enrollment, outcomes, and
benchmarking reports as each project
may require.
In addition, NERI statisticians employ
a wide range of advanced statistical
methods in analyzing data from registry
studies. Analysis of registry data can be
challenging due to incomplete or missing
data, potential confounders etc. Registry
statisticians have extensive experience in
managing these types of problems and
employ a range of statistical methods
and procedures to address these issues.
NERI offers a full range of registry services
including patient registries, post-approval
studies, post-market surveillance, patient
-reported outcomes, health resource utilization studies, CMS-mandated, patient
education, and market research.
11
NEW RESEARCH
A National Registry
to Help Understand
Women and Sexual Desire
E
xperiencing decreased sexual
desire can be bothersome and
distressing, so much so that it
has become an established medical
diagnosis called “hypoactive sexual
desire disorder” (HSDD). HSDD is very
common, affecting 10% or more of
adult women during the course of their
lives. The potential impact of HSDD
includes reduced quality of life, impaired
mental health, and problems with interpersonal relationships. Soon to be
published results from the Boston Area
Community Health (BACH) study show
that the problem affects women from
all three major race/ethnic groups about
equally.
Despite the prevalence and impact of
the problem, little is known about the
natural history of HSDD, such as how
it develops, progresses or resolves.
Screening and diagnostic tools for this
problem have not been evaluated in
large, representative groups of women.
While current treatment approaches
include hormonal, pharmacological, and
psychosocial therapies, research has yet
to describe exactly how these treatments
are used in actual practice, or what the
outcomes of different treatments tend to
be over time.
NERI, under the sponsorship of a private
pharmaceutical company, is taking the
lead to shed light on the mystery of
HSDD. To find out what HSDD means
in the real world, NERI is conducting a
multi-site, clinical registry study of
women experiencing this condition and
undergoing treatment in specialized or
general practice settings. The HSDD
Registry will include over 500 pre- and
post-menopausal women, tracking them
through time to follow changes in their
levels of sexual desire. In addition, the
study will assess other sexual functioning factors, physical and emotional
health, relationship factors, treatments
received, satisfaction with treatments,
and health care use.
“This study will break new ground in
our understanding of women’s sexual
problems and the role of current treatment practices,” according to Dr. Ray
Rosen, the Study PI.
Despite its prevalence
and impact, little
is known about the
natural history of
Hypoactive Sexual
Desire Disorder.
12
NEW RESEARCH
By providing a real-world view of clinical
practice and patient outcomes, registry
studies are increasingly becoming
appreciated for their ability to offer data
about disease impact and outcomes that
are not addressed by randomized trials
or traditional survey studies. For the
women distressed by HSDD, the Registry
may offer understanding and a better
sense of which treatments are likely to
help. This study will also advance scientific understanding of female sexuality
by establishing a national patient registry
for the most common sexual problem
in women.
An International Registry
to Explore Male
Androgen Deficiency
I
f a man’s testicles produce belownormal levels of male sex hormones
(androgens) he may be at risk for
a variety of health issues, ranging from
a reduced or absent sex drive to an
increased risk for prostate cancer.
Below-normal levels of sex hormones—
primarily testosterone—is termed
Androgen Deficiency or hypogonadism.
It may be caused by genetic malfunctions,
chronic illnesses (including HIV/AIDS),
or normal aging.
Population-based studies at NERI have
shown that about 1 in 5 men over the age
of 50 have abnormally low testosterone
levels. These men also have signs or
symptoms of hypogonadism, including
mood changes, irritability, sleep disturbance, and loss of sexual potency or
libido.
Despite the rapid increase in sales and
availability of both prescription and
non-prescription products for androgen
deficiency in men, few long-term studies
of the benefits or safety of hormonal
therapies have been performed. In addition, little is known about the natural
history and progression of male hypogonadism —whether the problem gets
better or worse over time (progression
or remission), and what effects androgen
deficiency in men may have on other
aspects of physical or mental health.
With funding from a private pharmaceutical company, NERI is conducting the
first large-scale, multi-national patient
registry study of male hypogonadism.
RHYME (Registry of Hypogonadism in
Men) is a long-term registry of men with
hypogonadism, which will include 500
men in North America and 500 men in
5 European countries who will be studied
for at least 2 years. Men who are receiving
various forms of hormonal therapy will be
included in the registry and compared
to men not on any treatment.
“This is the biggest current barrier to
men with hypogonadism seeking help
for their problem,” according to Dr. Ray
Rosen, the study PI. Recent studies have
suggested that, contrary to previous
beliefs, androgen deficiency itself may
be a risk factor for prostate cancer, and
this important question will be addressed.
RHYME will also investigate the longterm effects of hormonal therapy on
overall health and quality of life in carefully diagnosed and medically treated
hypogonadal men.
For the large population of men with
androgen deficiency, the RHYME
study will provide much-needed information about “real world” experiences
and outcomes of this common disorder
in aging men.
A major focus of the study will be on
the effects of hypogonadism and testosterone therapy on prostate health in
older men. In particular, does testosterone therapy increase the risk of
prostate cancer or benign prostatic
hypertrophy (BPH) in aging men?
13
ONGOING RESEARCH
Tracking Pain:
Does it Matter?
F
or the past decade NERI
researchers have looked
for ways to improve the
medical evaluation of patient
symptoms using digital technologies such as electronic note
pads and personal digital assistants (PDAs).
It’s been widely
reported that
patients prefer to keep
track of their own symptoms using electronic diary
tracking rather than pen-andpaper diaries. And the benefits
of electronic symptom tracking in terms of data quality are
well known. But does this kind of tracking actually improve
quality of care?
NERI recently completed a study that explored this question.
The research compared a PDA-based tracking system for
symptoms of chronic pain with a pen-and-paper system. The
study goal was to determine if the mode of tracking had any
effect on the actual experience of pain by the patients or to
the way the patients responded to pain in their lives.
A unique aspect of this study was the incorporation of qualitative interviews to gain insight into how users felt about the
type of tracking system they used and whether it helped or
hindered their pain management. Although previous quantitative results indicated that users felt it was easier to enter data
on the paper diary, the qualitative aspect of the study revealed
an interesting perspective on how patients and physicians
viewed electronic tracking and pain perception.
Physicians, although reluctant to add a new tool to their
already busy practice day, indicated that if such a tool were
part of their practice (rather than a study) it could be helpful
for tracking symptoms and understanding changes over time.
Patients similarly reported that being able to visualize changes
14
PHYSICIANS ON THE IMPORTANCE
OF TRACKING PAIN:
“I think it helps me better understand the
symptoms.”
“It gives me an idea over time of whether or not
the patient’s condition is improving, getting
worse, or staying the same.”
THE PATIENT PERSPECTIVE:
“It helped me face my pain a little bit more.”
“A lot of times I have trouble accepting my
pain. This makes me stop and at least be happy
that I can do what I can do.”
“I’m getting better at handling this…what used
to be a 9 I now consider a 6…”
over time helped them better understand their pain and
accept it as a part of their life they must learn to live with.
In particular, they commented that although their pain status
may not have changed, their approach to coping with it had.
Several papers are planned to report important findings from
this study, for example:
• Whether tracking impacts patients’ perception of care.
• What psychosocial components of tracking contribute to
pain management
• Does electronic tracking using the pain body drawing
improve patient’s or physician’s understanding of pain?
• How does reporting symptoms infrequently (monthly)
compare to daily data capture?
• Does electronic data tracking impact medication management?
This research is supported by NIDCR grant #: DE014797.
D I SC O V E R I N G R A C I A L D I F F E R E N C E S
I N M A L E B O N E ST R E N GT H
The NERI team of Andre Araujo,
Thomas Travison, Gretchen Esche, and
John McKinlay was joined by colleagues
Michael Holick and Tai Chen from
Boston University School of Medicine.
Participants were given bone scans and
were assessed for various measurements
of body size (height, weight, lean and
fat mass, etc.), gave blood for assessment
of bone-related measures (hormones,
calcium, vitamin D, etc.), and answered
questions about things related to bone
health (nutrition, fracture history, etc.).
The overarching aim of the study was
to obtain a better understanding of
osteoporosis and bone health in men,
particularly with respect to racial and
ethnic differences.
In the past 2 years, the NERI BACH/Bone
staff have published numerous papers
in top-notch peer-reviewed scientific
journals. We highlight some of the main
findings here.
Clarification of
the influence of lean vs.
fat mass on BMD.
Data clearly show that black men had
greater BMD than Hispanic and white
men. The differences between Hispanic
and white men were restricted to the
hip, and dependent on age. Age-related
declines in BMD were significantly steeper among Hispanic than among black
or white men, and this applied to almost
all skeletal sites in the data base. BMD
does not directly quantify bone strength,
however, since it measures material but
ignores aspects of the way that material
is arranged.
There is a long-running controversy
concerning the relative importance of
fat and lean tissue in stimulating bone
formation in adults. While many studies
suggest that fat mass is the prime determinant of BMD, others suggest that lean
mass plays a greater role. The BACH/Bone
data show that body mass index (BMI)
was associated with bone density only
up to certain thresholds (see Figure 2).
Thus, we examined age trends in hip
bone geometry, measures that consider
the arrangement of bone material. These
measurements were provided by Thomas
Beck from Johns Hopkins University.
Results revealed comparable age trends
in a measure of the bending strength of
bone, by race and ethnic group—results
which mirrored the BMD results (see
Figure 1).
Lean mass displayed strong and consistent associations. When the variable of
lean mass was statistically controlled,
the positive associations between bone
density and other body composition
measures (BMI, fat mass) disappeared.
(Results did not vary by race/ethnicity.)
The data therefore support the view that
it is the lean mass component of increased
body size that exerts a protective effect
on bone mass and that maintenance of
lean mass is the most promising strategy
continued on page 17
NARROW NECK
INTERTROCHANTER
1.15
1.15
Z (cm3)
Between 2002 and 2005, a study of
bone health was conducted among
male participants of the Boston Area
Community Health (BACH) Survey.
Called BACH/Bone, the study included
1,219 racially and ethnically diverse men
between the ages of 30 and 79 years.
Racial and ethnic differences
exist in bone mineral density
and bone geometry.
Z (cm3)
A
verage male bone mineral
density (BMD) varies significantly between racial and ethnic
groups, and these differences are not
due to differences in hormone levels or
measures of body composition, according
to groundbreaking new research by a
team of NERI investigators and their
collaborators.
1.00
0.90
1.00
0.90
0.75
0.75
30-39
50-59
70-79
30-39
Age (y)
Black Men
50-59
70-79
Age (y)
White Men
Hispanic Men
Figure 1: Proportionate age trends in geometric indicators of proximal femur strength in the
narrow neck and intertrochanter. Source: Travison et al., Osteoporos Intl, 2007.
15
ONGOING RESEARCH
Pediatric
Heart
Network
I
Childhood
Heart
Disease
An Ongoing
NERI Focus
n 2001, the Pediatric Heart
Network (PHN) was formed
to promote medical advances
in the treatment of childhood heart
diseases. Funded by the National
Heart, Lung, and Blood Institute, the
PHN is a group of eight hospitals in the
United States and Canada that conducts
research studies in children with congenital or acquired heart disease. NERI
is the data coordinating center for the
network. Led by Lynn Sleeper, NERI’s
PHN team has guided many aspects of
epidemiologic and clinical trial design
and implementation including:
• Statistical design and analysis
• Protocol development and execution
• Database management system
design and implementation
• Quality assurance
• Regulatory approvals and reporting
The PHN has three studies in analysis
phase, two studies completing subject
follow-up, two currently recruiting
patients, and one study set to launch
in mid-2008. (The studies completed
were described in detail in NERI’s 2007
newsletter.) In addition to conducting
research, the PHN takes seriously its
mission to be a resource to patients with
heart disease and families. NERI and the
16
NHLBI have jointly
developed a public
educational website
to provide information to families about
participating in clinical studies
and to describe ongoing PHN research:
www.pediatricheartnetwork.com. The
site offers a wealth of information that
helps parents with the sometimes-difficult
decision about whether or not to enroll
a child in a clinical study.
Marfan Syndrome: Finding Answers
The most recent PHN Trial, which began
enrollment in February 2007, is randomly
assigning patients aged 6 months to 25
years with Marfan Syndrome (a connective tissue disorder) to the angiotensin
receptor blocker losartan vs. the betablocker atenolol. Marfan Syndrome affects
many body systems including the eyes
and musculoskeletal system, in addition
to leading to dangerous enlargement of
the aorta. Recent animal studies have
shown dramatic prevention and regression of aortic dilation when losartan was
used. This FDA-regulated, Phase III
trial was launched by the PHN in direct
response to the promising animal studies
and has been featured in popular
magazines such as Readers’ Digest and
Women’s Day.
The site offers a
wealth of information
that helps parents with the
sometimes-difficult decision
about whether or not to
enroll a child in a
clinical study.
The trial will follow each patient for
three years. Ancillary studies involving
these patients, funded by the National
Marfan Foundation, are also underway,
and involve genetic analysis and a
detailed assessment of musculoskeletal
state and function. If successful, the new
treatment may delay aortic surgery and
provide hope to children and young
adults with Marfan Syndrome. As of late
February, the trial has enrolled 216 of
the target 604 patients at 19 pediatric
and adult cardiac care medical centers.
In 2008, eight additional medical centers
will be joining the trial in order to reach
target enrollment in 2010.
Single Ventricle Hearts
Normal hearts have four chambers: two
atria and two ventricles. Some children,
however, are born with only a single
functional ventricle, which greatly
reduces the efficiency of the heart. The
ongoing Single Ventricle Reconstruction
Trial is a landmark research effort in the
field of pediatric cardiology. Newborns
are randomly assigned to receive one of
two different types of shunt in the surgery, which must be performed within
days of birth to prevent death. The target
enrollment is 554 and as of late February
503 newborns have been enrolled, with
accrual expected to be completed by
mid-2008. Patient follow-up will continue
until mid-2009 to assess the trial’s
primary endpoint, death or heart transplantation by age one year.
Secondary endpoints of interest are
cardiac function, measured by echocardiographic imaging, and neurodevelopment. In addition to the question of survival, a compelling question
is how well these children can function
after an early childhood of hospitalizations and recovery—a critical developmental period in which their circulatory
system has been providing suboptimal
blood oxygenation. This trial is unique
because most other surgical studies have
not obtained long term follow-up of
these patients, and it is hypothesized
that the timing of adverse outcomes may
differ according to the type of shunt used.
Because the long-term follow-up of children born with a single ventricle is so
important, the PHN has developed and
received approval to conduct the Single
Ventricle Extension Study, which will
follow all survivors from the randomized
trial through their third stage of surgery,
called the Fontan procedure, which
typically occurs at age 2, and continue
contact and assessment of these children
up to at least 6 years of age. The study,
therefore, will continue until 2014. The
earliest enrolled child from the original
trial will be followed up to age 9.
The parents of the children participating
in the extension study will complete
annual assessments of their child’s abilities, behavior, and health-related quality
of life. Echocardiograms performed at
age six years will reveal how the heart is
functioning relative to earlier in life as
well as relative to normal children, and a
device to monitor heart rate and rhythm
will also be used for a 24-hour period
to capture information on possible
abnormalities such as heart arrhythmias,
which are of interest particularly in the
children who received one of the two
types of shunts. Lastly, a biologic specimen repository is being established to
store a blood sample
from participating
The Single
children as well
Ventricle
as their parents,
to be used in
Reconstruction Trial
future research
is a landmark
on the genetic
research effort.
bases of congenital
heart disease.
This research is supported by NHLBI
grant #: U01 HL68270.
DISCOVERI NG RACI AL DI FFERENCES I N MALE BONE ST RENGT H
continued from page 15
for preserving bone health with age.
Further analyses in the BACH/ Bone
cohort have shown that while black men
had the greatest total and proportionate
lean mass values, these differences in
body composition do not completely
explain racial and ethnic differences in
measures of bone density.
The role of hormones in bone density.
The relationship between hormones and
BMD in men has received considerable
attention. However, most studies have
been conducted in homogenous populations, and it is not known whether
differences in hormones impact race and
ethnic differences in BMD. We examined
Femoral Neck
8.0
8.0
6.0
6.0
4.0
4.0
3.0
3.0
15
25
35
BMI (kg/m2)
45
35 40
50
60
70 75
Lean Mass (kg)
Figure 2: Bone Mineral Strength in the Femoral Neck as it varies by Body Mass Index (BMI) and
Lean Body Mass. Source: Travison et al., Osteoporos Intl, 2007.
testosterone (T), estradiol (E2), and sex
hormone-binding globulin (SHBG) to
determine their association with BMD.
Evidence for race and ethnic variation in
the association between hormone levels
and BMD was very weak. We examined
25 interactions and found only one to be
significant. Total/free E2 levels were positively and significantly correlated with
hip BMD even with statistical adjustment
for age, lean mass, fat mass, physical activity, self-rated health, and smoking. However,
estradiol levels failed to account for race
and ethnic differences in hip BMD. These
data suggest that while E2 exhibits a modest
correlation with hip BMD, differences in
hormone levels are not a major contributor to the observed differences in BMD
between black, Hispanic, and white men.
This research is supported by NIA grant #:
BONE AG 20727.
17
ONGOING RESEARCH
TESTING A TREATMENT FOR
Heart Failure
W
hen a person’s heart is chronically weakened
and cannot pump hard enough to supply the
body with the oxygen and other nutrients it
needs, they are said to have heart failure. Heart failure is the
number-one discharge diagnosis in the US today. It accounts
for over 3 million hospitalizations annually—a figure likely
to rise as the population ages. NERI is the coordinating
center for a new clinical trial testing a drug treatment that
may significantly improve the health of a specific subset of
those with heart failure.
The left ventricle, one of the 4 chambers of the heart, pumps
oxygen-rich blood to the body. Each pump, or contraction,
“ejects” blood from the ventricle. A healthy left ventricle
typically ejects about 60% of its total blood volume with each
contraction. This is called the “ejection fraction.”
Studies have found that about half of patients with heart
failure have an impaired ejection fraction due to damage to
the left ventricle. A number of studies have focused on this
sub-population of those with heart failure. But the other half
of the population has intact left ventricle function and normal
or near-normal ejection fractions. Despite this, they, too,
suffer a range of problems and have a reduced survival rate.
This population has not been well-studied…until now.
A new clinical trial called TOPCAT is testing the effectiveness of a novel therapeutic approach to treating heart failure
patients with preserved ejection fraction. Specifically, the
study is looking at whether administration of spironolactone
(an aldosterone antagonist medication) will improve survival
and reduce hospitalization in the target population. TOPCAT
stands for Treatment Of Preserved Cardiac function heart
failure with an Aldosterone anTagonist.
Funded by the National Heart, Lung and Blood Institute,
the trial started enrolling patients in the summer of 2006.
It is a Phase 3 randomized, double-blind, placebo-controlled
study of 4,500 patients at about 200 U.S. and international
clinic sites. The study currently has subjects recruited in
North America (US, Canada), Latin America (Argentina) and
18
Eastern Europe (Republic of Georgia and Russia). In the
coming year, the trial will expand its reach to Brazil.
As of this writing, the study has already reached the milestone
of 1000 patients enrolled. The study is prospective and will
last about 4.5 years. It is one of the largest clinical trials of
heart failure ongoing today.
In addition to the primary objectives, the study will also determine whether treatment with the drug produces any clinically
meaningful reduction in the diagnosis of such conditions
as diabetes; atrial fibrillation; heart attack; stroke; kidney
impairment; or sudden death. If successful, the trial has the
potential to reduce individual risks of cardiovascular morbidity
and mortality in this understudied population, as well as have
major public health implications—reducing mortality and
costly hospitalizations.
The NERI team is led by Principal Investigator Rebecca Li
and co-PIs Drs. Marc Pfeffer (Brigham and Women’s Hospital)
and Sonja McKinlay. The team includes the following key
members: Carol Assang, Susan Assmann, Sue Blieden, Patrick
Compton, and Heather Cronin.
This research is supported by NIH grant #: N01-HC-45207.
Innovations in Clinical Trial Design
Adaptive Designs That Are Both Flexible and Rigorous
ADAPTIVE DESIGNS ARE GAINING
IN POPULARITY, BUT STILL NEED
RIGOROUS PLANNING.
continued from page 1
So how does it all work? When the trial
is well-advanced—typically when more
than half the patients have some outcome assessment and recruitment is
still ongoing—one performs an interim
analysis focusing on the current results
for the trial’s primary outcome.
An Expert Panel looks at the unblinded
results to see if they are sufficiently
favorable for the trial to have a good
hope of a positive result (i.e. new treatment wins!) upon trial completion at
the pre-agreed target sample size. This
requires calculating the conditional
power.
For example, if the observed results to
date were actually the true treatment
effect, what’s the probability of getting 5%
significance (P<.05) at the end? If the
conditional power is high (i.e. >90%)
then no problem, just proceed as planned.
But if it’s lower, you will need to increase
the trial size to improve your chances
of success. You’ll usually need to set
some realistic maximum on trial size for
reasons of cost and time.
So what’s the catch? Well, interim
results of ongoing trials are usually kept
top secret, so you need tight control
to ensure that only the Expert Panel
(and their assisting statistician) see the
unblinded data. Much recent statistical
research has been done on methods to
ensure that a trial’s type 1 error (risk of
any false positive claims) is not inflated
by “going adaptive.”
In a nutshell, such adaptive sample size
re-estimation requires rigorous planning
and documentation to make sure it
achieves its goals and is not a recipe
for a sloppy trial.
• Enrichment design (where you focus
in on a certain type [subgroup] of
patient that seem most responsive to
the new treatment)
• Play the winner rule (in which you
adjust the randomization ratios to
increase the proportion of patients on
the treatment that’s doing better to date)
As of this writing, however, the most
commonly used Adaptive Design is for
adjusting (increasing) the target number
of patients as you go along.
Adaptive designs have aroused suspicion
amongst clinical trial experts, who
suggest that such designs introduce an
element of inconsistency in approach.
But as Oscar Wilde said “Consistency is
the last refuge of the unimaginative.”
Other types of adaptive design include:
• Seamless Phase II/III trails (in which
you drop some arms in a multi-arm
trial, which may be useful in drug
dose selection)
19
NERI welcomes new staff
Jamie Bartkiewicz has
been hired as a Human
Resources Representative.
She previously worked
as a Human Resources
Coordinator at Children’s
Hospital. A graduate of St. Anselm College,
Jamie completed her BA in Psychology.
Megan Connor recently
completed her MPH
from Boston University
with an emphasis on
Epidemiology, and has
a BA in Biology from
the University of California/Davis. At
NERI she works as a Project Manager
on several private industry projects
with Ray Rosen. Prior to joining NERI,
Megan worked at Beth Israel Hospital
where she served as Project Manager in
the Urology Department.
Meghan Coughlin has
joined the Proposal
Development Department
in the role of Coordinator.
Meghan previously
worked at Brigham &
Women’s Hospital where she worked as
an Administrative Coordinator in the
Venous Thromboembolism Research
Group. Meghan received a BS in Biology
and Chinese from Union College.
Heather Cronin is now
working at NERI as
a Clinical Research
Associate on the
TOPCAT project. She
has an MSc, Environmental & Health, London School of
Hygiene and Tropical Medicine, and
a BS in Psychology and Biology from
Mary Washington College.
20
Anne DiLauro joins NERI
Chris Kenwood graduated
as an Associate Project
Director on the Pediatric
Heart Network. Anne
received her MPH
from the University
of Southern California and her BS in
Chemistry from Boston College. Anne
previously worked at the University
of Southern California as a Project
Specialist on a pediatric start-up project.
from UMass Amherst
with a MS in Biostatistics
and now works at
NERI as a statistician.
At UMass, Chris worked
on a graduate-school research project on
the quality of heath care with a professor
and physicians at Bay State Medical
Center.
Meena Doshi recently
completed her MPH
in epidemiology at
Tufts University. She also
has an MS in Applied
Statistics from the Ohio
State University, and a BS in Statistics
from the University of Pune, India.
Meena worked previously as a data
analyst and marketing analyst.
completed her PhD
in Biostatistics at the
University of North
Carolina, Chapel Hill
where she also earned
an MS in Biostatistics. Hae-Young
completed an MA and BA in Economics
at Ewha Womens University in Seoul,
Korea. She previously worked as a
statistical analyst at GlaxoSmithKline
and the CDC.
Emily Elstad. Prior to
joining NERI as Project
Manager on the Beneath
the Urologic Iceberg
study, Emily worked
as a Research Associate
for the National Center on Family
Homelessness. She completed her MPH
at Brown with a major in Qualitative
Research and International Health.
At Brown, Emily spent a term at the
Tafuna Family Health Center in
American Samoa as Field Director.
Hae-Young Kim recently
Additional New Hires
Sherani Amarasinghe
Clinical Research Associate
Hayley Arnett
Data Manager
Cheryl Browne
Associate Research Scientist
Michela Carlson
Receptionist/Office Assistant
Jennifer (Jenna) Erickson
Administrative Coordinator
Xiang (Bill) Kong
Eric Gerstenberger
completed his MS in
Statistics at UMass
Amherst, and has a BS
in Mathematics from
Drexel University.
Prior to joining NERI as an Associate
Research Scientist, Eric worked with
Fortelligent as a consultant/analyst for
a major wireless corporation.
Network Administrator
Zenaida Ramos
Bilingual Data Collector
Tehudis Salcedo
Data Collector/Field Tracker
Maria Sanchez
Data Collector
Glenda Singer
Senior Systems Analyst
LIFE AT NERI
NERI Presents at Drug Information
Association Meeting in June
NERI’S INFLUENCE
NERI scientists will chair two 90-minute speaking
reaching impact of NERI’s scientific
sessions and one poster presentation at the 44th
work, two papers published by NERI
Annual Drug Information Association (DIA) Meeting
staff are in the Top 10-most-frequently
In a vivid illustration of the wide-
in Boston on June 22-26. These NERI thought lead-
cited papers in the past 25 years on
ers will moderate and present in three different
the subject of sickle cell disease.
DIA tracks: statistics, clinical research and development, and regulatory affairs. NERI will also exhibit
at the conference and formally launch its specialized clinical research services to the private sector.
NERI President, Sonja McKinlay, will conduct a
discussion titled, Selecting the Optimal Sample
Size: Initial Realism or Adaptive Re-estimation.
In partnership with Stuart Pocock of the London
School of Hygiene and Tropical Medicine and Janet
A 1998 paper published in the journal
Blood titled “Cerebrovascular accidents
in sickle cell disease: rates and risk
factors,” was the 2nd-most-frequently
cited paper.
A 2000 paper published in the New England
Journal of Medicine titled “Predicting adverse
outcomes in children with sickle cell disease,” was
the 6th-most-frequently cited sickle cell reference.
Wittes, Statistics Collaborative, Washington DC,
McKinlay will explain ways to estimate appropriate
sample size, as a key factor in optimizing design
efficiency (and ultimately minimizing study cost).
Rebecca Li, NERI’s Vice President of Clinical Research,
will lead a session titled, Forging Partnerships
with NIH-sponsored Clinical Trials and Networks.
In collaboration with Michael Domanski of the
National Heart Lung and Blood Institute and Dean
C. Winter, Vice President of Scientific and Clinical
Affairs for AtCor Medical Inc., Li will provide attendees with a perspective on how partnering with
the NIH can be a cost-effective path to furthering
drug and device development in the private sector.
Leveraging NIH-sponsored clinical trials and networks is an underutilized tool for biotechnology
and pharmaceutical companies.
Finally, Flora Sandra Siami, NERI’s Director of
Regulatory Affairs, will present a poster titled,
Ensuring Compliance and Quality in Global Clinical
Giving Back to the Community
In 1999 NERI employees started what
has now become an annual tradition:
the “Snow Kids” project. Each year
NERI contacts the local Department of
Social Services and is given information
about families in need. This year NERI
“sponsored” 5 families including: a single
mom with a 4 year old child; a single mom and teenage daughter who
had been homeless due to a job loss; and a family with 6 kids struggling
to make ends meet.
At holiday time, NERI staff sign up to buy something from the “wish list”
of gender and age-appropriate gifts for each family member. Present
requests usually include both utilitarian things like pots, pans, dishes and
warm boots, as well as fun things for the kids, such as IPods or bikes.
This year NERI staff gathered about 130 presents and gift certificates—
and DSS had to send three vehicles to collect it all!
Trials. The poster will provide attendees with
diagnostic tools that can be incorporated into the
management of global clinical trials. Attendees
will be able to better evaluate the compliance and
adherence of an international clinical research site.
In a Thank-You letter sent to NERI, DSS wrote: “We only wish that you
could see for yourself how gratefully each gift was received. Your
support provided to families over the holidays not only helps to ease
their financial burden but reassures them that many people in the
community care about their wellbeing.”
21
LIFE AT NERI
Administrative
Coordinators new
to NERI
From left: Amanda
(Mandi) Cass,
Anne (Patti) Muehter,
Diane Barrios
New Team
Members
Three of the largest groups of employees
at NERI are Administrative Coordinators,
Data Collectors, and Programmers/Data
Managers. The dedication to quality work
that these teams exhibit is one reason for
NERI’s sterling international reputation for
the development and implementation of
clinical trials and other kinds of scientific
research projects.
Data Collectors new to NERI
From left (front): Journel (Jean)
Joseph, Luz Morillo, Felecia
McCray. From left (back):
Manuel Burnias, Maria Carrera
Programmers and Data
Managers new to NERI
From left: Patrick Compton,
Nathan Li, Elliot FletcherSananikone, Jessica Brewer,
Shaili Puligari
22
LIFE AT NERI
Data Management at NERI
N
eri’s web-based ADEPT system
provides a secure, cutting-edge
platform for all data-management
activities. It is optimized for multi-site,
international studies, greatly simplifying
data management.
Maintaining ADEPT’s position as a
state-of-the-art tool requires constant
improvements. To that end, several
features have recently been released:
• Case Report Form (CRF)
Authoring Tool
• Online CRF Library
• Study Metrics Report Generator
• Updated look and feel
The CRF Authoring Tool will allow data
managers to develop Case Report Form
content, specifications and instructions
concurrently. An integrated environment
will allow immediate testing of data
items, validations, skips, and help text.
ADEPT’s online CRF Library contains
more than a hundred CRFs and metadata, all categorized by subject. Data
managers can browse by category
and select either an entire previously
validated CRF or just sections of interest.
This CRF library is accessible from
within the new CRF Authoring Tool,
enabling very rapid CRF development.
A study metrics report generator allows
data managers to define custom reports
built upon a core set of study metrics.
These metrics cover CRF data entry
status, QC double data entry, late and
early visits, electronic signatures and
basic accrual. In addition to an initial
20 standard metrics, calculated metrics
can be derived from existing metrics.
It is also possible to easily add project
specific metrics. Any metric report can
be described by up to 3 graphical locations (i.e. continent, country and region)
as well as site, study event, and or CRF.
In addition to these significant improvements, all ADEPT reports have a new
look and feel including new icons, and
a simpler, cleaner table layout.
Keepin’ It Green at NERI
Here are a few of the things NERI is doing to reduce its “carbon footprint”
and conserve natural resources:
• Recycle all paper and paper products
• Recycle printer cartridges
• Recycle old computers and printers
• Encourage employees to use public transportation by providing discounts
on MBTA passes
• Use the “energy saver” mode on all copiers and fax machines
23
Organized Around Excellence
NERI is a plural noun.
NERI is not an “institute” but a group of institutes, each dedicated to
a specific area of scientific expertise, such as clinical trials, gerontology,
or epidemiology. Each institute is staffed by a wide variety of research
scientists: biostatisticians, statisticians, epidemiologists, clinical research
monitors, data managers, programmers, network analysts, and other
workers with advanced degrees in scientific disciplines.
Although not obvious to those outside the company, the institute
structure at NERI provides many benefits, both to NERI itself and its
clients. The institutes allow NERI to provide a full service approach
to clinical trials (both NIH and private industry sponsored), social and
public policy research, and medical and policy studies of a wide range
of patient populations. NERI’s world-renowned scientists are available
across all projects to provide consultation and expertise. In addition
to the institutes, NERI has established centers for statistical analysis
and study design, data management, and data technology that provide
service across projects and institutes. This allows NERI to be nimble
in shifting priorities, and cross-training staff to meet the demands of
an ever-expanding client base.
The company was purposely designed to promote this sharing of knowledge and expertise. Not only does it provide benefit to our customers,
but also provides a stimulating and challenging work environment for
NERI employees to expand and develop a wide knowledge base.
PRSRT STD
U.S. POSTAGE PAID
PERMIT #56435
WATERTOWN, MA
02472
9 Galen Street
Watertown, MA 02472
Please send all
address corrections to
[email protected]
`