[Excerpt : IHI ADMINISTRATION Manual - Volume 2, pp56-58]

Schedule 12: Data Centralisation, Archiving and Data Sharing

IHI aims at providing more value to its stakeholders and increase the efficiency of information generation, synthesis and dissemination. IHI also wants to participate in the global efforts in making more information available in the public domain so more scientists and investigators can make use of this resource.

Most of the initial information is collected in collaboration with the IRB office. Follows a close collaboration beteween the study/project and IHI's librarian.
DDI standard and the NADA system are the bulk of the archiving procedure.
For any question related the procedure, please refer to the "Contacts" page .

To increase efficiency of data collection IHI will continuously upgrade it systems for data collection and collation using point of observation electronic data capture (e.g. PDA, Tablets), automatic and rapid transfer (automatic synchronization with GSM, G3) and automated data checks and simple analytical routines. Satellite servers will be used for the different activities/projects (also in different geographical locations/branches)and these will feed into a centralized secure data servers. The centralization will be facilitated by the ongoing efforts in using similar software, SOPs and data structures across the institute for similar work.

• Data centralization will involve collection of data in central electronic bank (Central Servers) and storage of this data in an electronic Archive. Both legacy data and data from ongoing projects will be centralized.

• Data centralization will apply for all projects and activities of IHI, both scientific and administrative.

• Centralization and central archive is mandatory for all data collected in the frame of IHI contracts and institutional programmes.

• Data from students projects funded or supported by IHI research infrastructure will also be centralized.

• Data collected as part of national activities (e.g. Health Management Information System) will be centralized and mirrored to or from appropriate National level hosts servers. Specific data use restrictions if any, will be part of metadata descriptions.

• Standard Metadata (data about data) form will be completed by data generators and made available to the Central Server Administrator.

• Progress reports of projects and activities that are provided by project or activity leaders twice a year will also be compiled through the portfolio manager. An intra-net progress report form will be developed and used. The summary of the progress reports will be generated from the Central Server.

• Electronic sign off from project/activity leader is required for documents and data to be uploaded into the centralized database.

• Continuous data update (daily) to the Central Server of ongoing activities will be done with facilitation of the ICT department.

• Automated checks at central level will be done routinely by the Central Server Administrator and feedback will be provided to activities leaders for database corrections and updates.

• Automatic database lock will be done after one year from the last data collection event in a time limited project or activity or one year after the calendar year of data collection for continuous (e.g. Sentinel Panel of Districts) projects/activities. Notification of data base lock after completion of data collection and checking will provided to the activity leaders and indicated in the intra-net.

• Databases in the Central Server will have accompanying data dictionaries provided by activity leaders or their designate.

• Repository of electronic copy of the investigator file (including protocols, informed consent form template, data collection tools, approvals, Sponsor agreements, Material transfer Agreements etc) will be done. If there are any restrictions on future use of data generated this will be indicated in the metadata.

• Original data sets (e.g. harmonized variables in centralized data bases for HDSS from Kigoma, Rufiji, Ifakara) will be archived in the Central Server Archive.

• Public access to the data will be provided in several formats for unrestricted databases and hiding confidential information (e.g. Names). Unless strong justification is provided, only commercial entities sponsoring activities of IHI will be allowed to put restrictions on public access of the databases. IHI will strive to have most of its databases unrestricted especially those funded by commercial companies.

• A public support center at IHI will provide additional information about the databases and address specific queries from the public.

• Public access approval system for Central Server Databases

• All administrative databases will be permanently restricted.

• The project/activity specific database will be exclusively available to project or activity specific Investigators during data collection and for 1 yrs after the database lock. During this period the databases will only be advertised in the IHI intra-net and be labelled restricted or not available.

• After the first two years above, access to the databases will be exclusive to Analysis cluster or other IHI scientist or project Collaborating partners in consortia/multicenter studies or analytic projects with capacity building for IHIfor a period of up to 2 yrs. During this period the databases will be advertised in the general public domain but labelled restricted and not available.

• Three years after data base lock the data sets will be advertised on IHI website and availability for download indicated. Download approval will be required from the Scientific Committee and IHI IRB where approval is indicated. In cases of disputes the IHI Director’s decision is final. Download of data will also require agreement from the recipient of the terms of data material transfer agreements as prescribed by the National Medical Research Coordinating Committee (MRCC) and Tanzania Commission for Science and Technology (COSTECH) and any other body specified by the Government of Tanzania. Public access of new data will be made on an annual basis.

• The specifics of databases for public access are as follows:

• Non searchable databases: Listings only and metadata of the databases will be viewable to the public. The public can request for simple and complex analysis to be done by IHI data cluster through the IHI support center. Download option will be available for non-restricted databases. Download approval is required.

• Searchable data sets: Listing, metadata, searchable functionality and simple statistics on web fronts will be available. The public can request complex analysis to be done by IHI data cluster through the IHI support center. Download option will be available for non-restricted databases. Download approval is required.

• Publicly available data sets (e.g. Non Human Subject, non IHI data sets with no restrictions etc) : listing, metadata, searchable functionality and simple statistics on web fronts will be available. The public can request complex analysis to be done by IHI data cluster through the IHI support centre. Download option will be available. Download approval is NOT required.

© IFAKARA Health Institute (IHI), 2017