Article type
Year
Abstract
Background: The Systematic Review Data Repository (SRDR) is an example of an open-access, state-of-the-art, web-based data system that allows online form building, data entry by multiple users, and data management and sharing.
Objectives: To describe our experience using SRDR for data abstraction.
Methods: The project leader creates the project and abstraction forms in SRDR following user-friendly wizard. S/he then adds users to the project and assigns roles. Users extract and enter data directly from study reports into SRDR online for seamless data saving and procession. The Data Comparison Tool within SRDR allows identification and adjudication of discrepancies. SRDR can be used to keep track of progress with data collection and workflow. The data can be exported out of SRDR into an analyzable format or a report. SRDR allows different alternatives for archiving and publishing the data.
Results: The Box summarizes steps for developing abstraction forms in SRDR. We successfully used SRDR to collect data from about 400 trials involving 15 data abstractors (800 records in SRDR due to double data abstraction). Our form contains 125 questions and 550 data items. It takes on average an hour for one data abstractor to collect data from one article and another hour for adjudication of discrepancies. Because the data can be directly exported in an analysis-ready format, we have used SAS c for data management and analysis. Because SRDR codes and stores data in distinct data fields, evidence tables and risk of bias tables can be easily generated.
Conclusions: SRDR has facilitated efficient data collection and preparation of a dataset from a large number of studies. To enable authors of Cochrane systematic reviews to use this freely available tool for data collection, future efforts must focus on developing tools to link SRDR to RevMan.
Objectives: To describe our experience using SRDR for data abstraction.
Methods: The project leader creates the project and abstraction forms in SRDR following user-friendly wizard. S/he then adds users to the project and assigns roles. Users extract and enter data directly from study reports into SRDR online for seamless data saving and procession. The Data Comparison Tool within SRDR allows identification and adjudication of discrepancies. SRDR can be used to keep track of progress with data collection and workflow. The data can be exported out of SRDR into an analyzable format or a report. SRDR allows different alternatives for archiving and publishing the data.
Results: The Box summarizes steps for developing abstraction forms in SRDR. We successfully used SRDR to collect data from about 400 trials involving 15 data abstractors (800 records in SRDR due to double data abstraction). Our form contains 125 questions and 550 data items. It takes on average an hour for one data abstractor to collect data from one article and another hour for adjudication of discrepancies. Because the data can be directly exported in an analysis-ready format, we have used SAS c for data management and analysis. Because SRDR codes and stores data in distinct data fields, evidence tables and risk of bias tables can be easily generated.
Conclusions: SRDR has facilitated efficient data collection and preparation of a dataset from a large number of studies. To enable authors of Cochrane systematic reviews to use this freely available tool for data collection, future efforts must focus on developing tools to link SRDR to RevMan.