the bronchiectasis research registry: a resource for … · the bronchiectasis research registry...
TRANSCRIPT
Screen Shots: Custom reports - Modeling
Screen Shots: Custom Reports
The Bronchiectasis Research Registry: A Resource for Collaborative Research in non-Cystic Fibrosis BronchiectasisL. LaVange, D. Stewart, B. Thomashow, K. Olivier, M. Knowles, C. Daley, and A. Barker for the Bronchiectasis Research Consortium
Introduction Results Screen Shots: Standard reports – Univariate Analysis
Bronchiectasis is characterized by inflamed bronchi that are abnormally dilated and chronically infected. Significant morbidity and mortality are associated with this progressive disease that afflicts thousands of patients.
The Bronchiectasis Research Registry was designed to support collaborative research and assist in the planning of future therapeutic clinical trials with the over-arching goals of improving patient management and furthering our understanding of the diagnosis and pathophysiology of the disease.
Methods
The Bronchiectasis Research Registry is a consolidated database of non-cystic fibrosis patients from major clinical and research institutions (n=10) across the US. The Registry was establishedby the Bronchiectasis Research Consortium, is funded by the COPD Foundation, and is governed by the Registry Advisory Board.
The Coordinating Center (CC), located at the University of NorthCarolina in Chapel Hill, developed and maintains a secure, web-based data management (DMS) and interactive reporting system for the collection, processing, storage, and analysis of data.
Adult bronchiectasis patients are asked to participate and consent to (1) use of their data for research purposes and (2) future contacts concerning participation in clinical trials.
Data acquisition is conducted through medical records abstraction and/or clinic examination and includes demographics;medical history; and clinical procedures pertinent to the treatment of bronchiectasis, namely respiratory symptoms, pulmonary function testing, lung imaging, diagnostic tests, and therapies.
The interactive reporting system provides Consortium investigators with user-friendly, real-time access (24-hour currency) to both standard and custom analyses of Registry data.
DMS
The Registry Data Management System (DMS) is a full service, web-based system providing all the capabilities required for research data management:
• Data entry, transfer, and validation• Database updating, retrieval, and closure• Data confidentiality, security, and archiving • Local (center specific) reports on questionable values and data
completeness• IT Security Plan in accordance with NIH IT Systems Security
requirements
The Interactive Reporting System is a web-based system accessed through the DMS, utilizing all the security features build into the DMS and providing:
• Reports based on data collected through the previous day• Standard reports generated nightly in PDF form, including
• Univariate distributions (tables and graphs)• Summary tables for related groups of variables• Results overall and stratified by gender and site
• Custom reports generated in real-time based on user requests• SAS® programs run remotely on a secure project server• Results returned to the user’s web browser• Automatic error checking of user requests (e.g., only
categorical variables can be selected for cross-tabulation)• Reports requested for all Registry patients or for the subset
consenting to be contacted for future studies
Data collection began in Fall 2008. To date, 500 patients have been enrolled from 9 centers, and 93% have consented to future contacts. Ninety-two % of expected data collection forms have been entered into the DMS, with complete data obtained for 78% of participants. Annual follow-up data collection began in Spring 2010 to update contact information and clinical data, including health status, clinical endpoints, respiratory symptoms, therapies, clinical procedures (lung functions) and laboratory microbiology.
Future plans include expanding recruitment to additional clinical centers, linking to stored image data for central reading, and expanding the DMS to incorporate the transfer of existing electronic medical records from clinical centers.
Stratified Analysis
Standard reports: Univariate
Summary Table
Model SelectionExample: Model Selection
Descriptive Statistics
Cross-Tabulation
Model Results
Format Optimized for Printing
Graphical Presentation
Scatter Plot Reflecting Model
Interactive Reporting System
Future Plans