Data storage and security

The database will be hosted on a web-based application designed to allow researchers to upload, view and manage data.

An Open Data Kit (ODK) platform(32) will be deployed for this purpose. Community and hospital survey data will be collected on ODK-programmed tablets and uploaded via secure connection to servers at KCRI.

The ODK platform will be used to produce a study-specific application and shall provide interoperability functions, such as exporting data to excel spreadsheet and other statistical packages.

This database will be developed and approved by the Data Management team for the study before utilisation. The data will be stored on 3 servers: primary, mirror and backup.

The primary server will be used to process incoming ODK, before being backed up on mirror and backup servers. All servers are behind firewalls and locked in secure cabinets.

All quantitative data will be initially stored within KCRI servers in the ODK platform. On completion of the study, or as part of routine data monitoring, the data will be extracted from ODK for analysis using statistical packages such as STATA, SPSS, R or SAS.

All content analysis data will be stored as Excel spreadsheet files and transcripts stored as Microsoft Word files. Final versions of all datasets and documents will also be exported to, and made available, as ASCII and/or CSV data files, with accompanying command/syntax files, so future users will still be able to access the data.

The audio data collected during qualitative data collection will be saved separately with no participant identification information.

Each data file will be catalogued in a single database, with accompanying metadata (e.g., filename, author, abstract, producer, geographic coverage, temporal period of collection, response rate, etc.) using Data Documentation (DDI) Initiative standards.