Article type
Year
Abstract
Introduction: In 1996, the U.S. Center for Disease Control and Prevention (CDC) responded to the needs of prevention providers and researchers by initiating the HIV Prevention Research Synthesis (PRS) project to develop an ongoing database of HIV prevention studies for research synthesis activities. The Studies Tracking, Acquisition, Review, and Synthesis (STARS) system was developed to support this project and serve as a cumulative, central repository of HIV prevention literature. The resultant database of trials is one of the primary sources of trials for the Cochrane Collaborative Review Group (CRG) on HIV infection and AIDS Prevention Registry
Objectives: To describe the development and core features of STARS system, including linkages with the research synthesis process and the temporal updating of ongoing systematic reviews
Methods: The database system was developed to parallel the systematic review process and to collect data central to critical phases of the research synthesis process (e.g., fugitive literature tracking, data evaluation, etc.). Relational database were developed in Microsoft Access. Electronic data entry forms were created for each database and internal quality control checks were built into the system. Report generation features were added to generate reference lists, frequency distributions and exportable data files for analysis with SAS or SPSS.
Results: The development of the STARS system yielded four linked primary databases: (1) Fugitive Literature Tracking database, which contains records for "hard-to-find" literature or newly funded research projects; (2) Bibliographic database, which compiles the results of manual and automated searches and includes key bibliographic information for each record; (3) Data Evaluation database, that contains the results of the study selection process (i.e., screening for relevance and methodological rigor); and (4) Analysis database, that consists of data and statistical information used to calculate effect sizes for specific outcomes. Automated database quality control checks improved data abstraction accuracy.
Discussion: Almost three years of development resulted in a comprehensive database system to support the systematic review process, including reviews generated by the prevention track of the Cochrane CRG on HIV infection and AIDS. Further development of the STARS system will maximize its dissemination an coordination capabilities and its generalizability to other science-based systematic review topics. Moreover, the ongoing, cumulative nature of this database system will facilitate updating systematic reviews with new evidence in a timely fashion.
Objectives: To describe the development and core features of STARS system, including linkages with the research synthesis process and the temporal updating of ongoing systematic reviews
Methods: The database system was developed to parallel the systematic review process and to collect data central to critical phases of the research synthesis process (e.g., fugitive literature tracking, data evaluation, etc.). Relational database were developed in Microsoft Access. Electronic data entry forms were created for each database and internal quality control checks were built into the system. Report generation features were added to generate reference lists, frequency distributions and exportable data files for analysis with SAS or SPSS.
Results: The development of the STARS system yielded four linked primary databases: (1) Fugitive Literature Tracking database, which contains records for "hard-to-find" literature or newly funded research projects; (2) Bibliographic database, which compiles the results of manual and automated searches and includes key bibliographic information for each record; (3) Data Evaluation database, that contains the results of the study selection process (i.e., screening for relevance and methodological rigor); and (4) Analysis database, that consists of data and statistical information used to calculate effect sizes for specific outcomes. Automated database quality control checks improved data abstraction accuracy.
Discussion: Almost three years of development resulted in a comprehensive database system to support the systematic review process, including reviews generated by the prevention track of the Cochrane CRG on HIV infection and AIDS. Further development of the STARS system will maximize its dissemination an coordination capabilities and its generalizability to other science-based systematic review topics. Moreover, the ongoing, cumulative nature of this database system will facilitate updating systematic reviews with new evidence in a timely fashion.