This is a model of the electronic database that we will be using in
this study if this study is funded. This database will closed and password
protected prior to initiation to the trial.
Database for Genetic Determinants of PAD
Data Systems and Management
Creation of the database infrastructure
The Biostatistics core of this project is led by Dr. Iain Johnstone,
Professor and former Chairman of the Stanford University Dept of Statistics. This
project will require several databases that will need to be created, subjected
to quality control measures, stored, protected against loss and intrusion,
and linked for subsequent analysis. These databases include the results
of chart review of cases and controls, the results of patient interviews
and exams, blood tests, vascular compliance measures and genetic
data.
Data from the chart review and patient examinations will be
entered by the nurse-researchers onto secure online electronic forms.
The data goes directly into a central database. Patients will be identified
by a unique study identifier, and a separately stored, independent translation
table will allow linkage of research data to the medical database of
the individual patients for longitudinal study. Genetic data will be
generated by Dr. Quertermous' group at the Stanford CVRC using coded
samples provided by Mt. Sinai and Stanford. These samples will be identified
using only the subject's unique study identifier, so that the CVRC personnel
will be blinded as to the identity of the subject, their clinical data,
and their status as a case or control. Results of the genotypic analysis
for each patient will be stored in a computerized database, and subsequently
files will be exported on a monthly basis to the Biostatistical Core
for analysis under the direction of Dr. Iain Johnstone. The computer
databases at the CVRC will be backed up daily, and archival copies will
be created monthly for off-site storage.
Data analysis
The clinical datasets and genetic datasets will each be identified
by the same unique, encrypted study identifier, and these datasets will
be merged at the Biostatistical Core Lab to form an analysis dataset.
The Biostatistical Core Lab will check the data for completeness and
consistency, working with the research nurses at Mt. Sinai and Stanford,
and the CVRC as needed to resolve any apparent inconsistencies or data
errors. All data is stored on a server in the Division of Biostatistics
in Oracle datasets and backed up on a daily basis. Archival snapshots
of the database will be created at regular intervals and stored offsite
in the event of damage to computers or building (e.g., fire, earthquake).
)