Article type
Year
Abstract
Background: The process of performing a systematic review and meta-analysis is complex. Usually, several people are involved and large amounts of data have to be managed. Commercial software is available to manage the process of performing a systematic review but it is expensive.
Objectives: To develop a data management software for systematic reviews and meta-analyses. In detail, the system should be able to facilitate the process of literature screening, data extraction, quality assessment, and meta-analysing data. The system should be available in-house as well as from remote locations.
Methods: Using Microsoft Structured Query Language (MS SQL) as backbone and Microsoft Access as a frontend we developed a data management software. The database is stored on a central server located at our institute. The Remote Desktop Protocol is used to provide terminal services for remote database usage. Meta-analysis are done using Stata and the application programming interface Open Database Connectivity (ODBC) provides direct access to all data via the statistical package.
Results: So far, we have been able to implement a comprehensive data extraction tool. All data relevant in the conduct of a systematic review can be extracted except time-to-event data. Data extraction forms are not flexible but implemented as generic forms. The tool is able to manage the consensus process considered standard in the conduct of systematic reviews. Only data provided in a study report needs to be extracted. No calculations by hand are required. Instead, all calculations are done by a validated program in Stata (e.g. calculation of effect sizes from means and standard deviations). The direct import of all data into the statistical package Stata avoids typing and transcription errors.
Conclusions: We plan to further extend the system. In the future we will allow for extracting time-to-event data and coverage of literature screening which will also provide management of references. Furthermore, flexible data extraction forms will be developed with the use of Microsoft C. However, to ensure further development of the system we need additional funding. To this end, we may also provide the system based on open source software.
Objectives: To develop a data management software for systematic reviews and meta-analyses. In detail, the system should be able to facilitate the process of literature screening, data extraction, quality assessment, and meta-analysing data. The system should be available in-house as well as from remote locations.
Methods: Using Microsoft Structured Query Language (MS SQL) as backbone and Microsoft Access as a frontend we developed a data management software. The database is stored on a central server located at our institute. The Remote Desktop Protocol is used to provide terminal services for remote database usage. Meta-analysis are done using Stata and the application programming interface Open Database Connectivity (ODBC) provides direct access to all data via the statistical package.
Results: So far, we have been able to implement a comprehensive data extraction tool. All data relevant in the conduct of a systematic review can be extracted except time-to-event data. Data extraction forms are not flexible but implemented as generic forms. The tool is able to manage the consensus process considered standard in the conduct of systematic reviews. Only data provided in a study report needs to be extracted. No calculations by hand are required. Instead, all calculations are done by a validated program in Stata (e.g. calculation of effect sizes from means and standard deviations). The direct import of all data into the statistical package Stata avoids typing and transcription errors.
Conclusions: We plan to further extend the system. In the future we will allow for extracting time-to-event data and coverage of literature screening which will also provide management of references. Furthermore, flexible data extraction forms will be developed with the use of Microsoft C. However, to ensure further development of the system we need additional funding. To this end, we may also provide the system based on open source software.