|
|
Home / News &
Events / Pre-Analytic and Post-Analytic Factors in Postmarket Drug Safety Data Mining
Pre-Analytic and Post-Analytic Factors in Postmarket Drug Safety Data Mining
Presented on January 24, 2006
Speakers: Alan Hochberg, Vice President of Research,
ProSanos Corporation, and Michael O'Connell, Director
of Life Sciences, Insightful Corporation
Listen
to the archived webcast.
Helpful Links:
The pharmaceutical industry is now responding
to a series of high-profile drug safety issues in a number
of ways, including the development and deployment of new
methods of visualizing and mining drug safety information.
Many of these methods employ state-of-the-art techniques
in statistical analysis and graphical display. In developing
and deploying these methods, it must be kept in mind that
pharmacovigilance is a man-machine partnership: a computer
is used to collect, digest, and display large amounts of
data regarding drugs and adverse events, while a human drug
safety expert, trained in biochemistry, physiology, and
medicine as well as statistics, interprets the data and
makes decisions. This differs from other applications of
data mining, such as in some areas of finance, where decisions
are made by automated systems themselves.
The quality of the output from the man-machine drug safety
partnership is determined by a large number of pre-analytic
and post-analytic factors, as well as the quality of the
analytics themselves. Data quality is a major determinant
of the effectiveness of pharmacovigilance. Issues in the
input data stream include missing data (which may be disguised
in various ways), duplicate records, coding and classification
errors, and inconsistent spelling. Phenomena other than
drug adverse events are often reported, while actual adverse
events are missed. Publicity or legal action can bring
a flurry of adverse event reports, complicating efforts
to use data mining algorithms that depend on temporal
factors.
Once a drug safety signal is generated, a number of post-analytic
factors come into play. Most of these are a consequence
of the fact that postmarket drug safety monitoring does
not provide the kind of information regarding causal relationships
that comes from randomized, controlled clinical trials.
An appreciation of the pre-analytic and post-analytic
factors in drug safety should be helpful to anyone who
uses statistical or graphical methods for the analysis
of postmarket drug safety data, especially when adopting
the new methods of statistical data mining and graphical
analysis as an adjunct to traditional clinically-oriented
methods of pharmacovigilance.
Alan Hochberg, Vice President
of Research, ProSanos Corporation |
Alan
Hochberg is Vice President of Research for ProSanos
Corporation in Harrisburg, PA. He and his group focus
on the development of new statistical methods for
the pharmaceutical industry, including pharmacovigilance
methods and disease progression models. Mr. Hochberg
has worked as a technical professional in bioinformatics
and biomedical engineering for the past 25 years.
He joined ProSanos from DuPont Qualicon, where he
was involved in the development of analysis software
for bacterial DNA fingerprinting, in the creation
of knowledge bases for infectious disease epidemiology,
and in the development of automated DNA sequencer
software. Previously, he worked for Hologic, Inc.,
Johnson & Johnson, and American Science &
Engineering, as a software developer, instrumentation
engineer, and manager. Mr. Hochberg holds several
patents on optical immunoassay devices, and on data
analysis methods for biologically-derived signals.
He received his B.S. degree in Electrical Engineering
from Princeton University. |
Michael O'Connell, Director
of Life Sciences,
Insightful Corporation
|
Michael
O'Connell has been working in the medical device,
diagnostics, pharmaceutical and biotech arena for
the past 15 years. Dr. O'Connell's background and
graduate work was in applied statistics and he has
published more than 40 papers on statistical methods
and life science applications including calibration,
mixed models, and nonparametric regression. He has
also written several statistical software packages
and libraries using S-PLUS, R and SAS. Most recently
he has been active in bioinformatics and the statistical
analysis of microarray data; and in the development
of tools for analysis and reporting of clinical and
safety data from S-PLUS.
Dr. O'Connell holds a Bachelors degree in Science
from the University of Sydney, a Masters degree
in Statistics from the University of New South Wales
and a Ph.D. in Statistics from North Carolina State
University. |
|
