B. Data: management, integrity, retention and preservation
B.1 MRC-funded research covers a broad spectrum, including fundamental lab-based science, population health science, patient-based studies and clinical trials. All research data generated through MRC-funded research must be managed and curated effectively throughout its life-cycle, including archiving, to ensure integrity, security and quality and where possible to support new research and research data sharing to maximise the benefit and impact of MRC research funding. Records should be kept to enable understanding of what was done, how and why, and which allow the work to be assessed retrospectively and repeated if necessary.
B.2 The principles, standards and technical processes for data management, retention and preservation will be determined by the area of science, for instance, the needs of fundamental studies may differ from population-based and clinical studies. Processes should be supported by appropriate data standards addressing confidentiality and information security, monitoring and quality assurance, data recovery and data management reviews where suitable.
B.3 Research data generated in the course of MRC-funded research must be retained in an appropriate format in line with agreed retention periods (see Box 1 below). Related material should also be retained to ensure accessibility, integrity, the application of new analytical tools or statistical techniques, and to enable new research where possible (see Box 1 below):
Box 1: Retaining research data
Types of research data
Primary/raw data include any information, images, observations, questionnaires, products, devices, procedures and any other data sources that are generated, recorded or used during the project.
Related material includes approvals, information on the consent process (including signed consent forms), meta-data, information relating to analysis and methodology, quality assurance records that demonstrate the validity of the data and/or adherence to experimental protocols, calibration data, collaborative agreements, intellectual property ownership, management and agreements and other relevant correspondence.
Confidential identifiable data comprise any information that relates to an individual (either living or dead) from which that individual can be identified.
The main record refers to the record where the research data are primarily recorded (for example, in electronic or hard copy laboratory notebooks, clinical trial master files).
Requirements
Retention periods Retention periods for primary/raw data and related material should be considered at the outset and should reflect any legal and regulatory requirements and, where possible, the aim to support new research.
The MRC’s expectations for research data retention are: Basic research
Population health and clinical studies
Where complete retention of all research data is not appropriate, the data must be validated using quality assurance procedures and the justification carefully assessed before any data are destroyed. Further guidance relating to retention periods can be found in the Clinical Trials and Data and Tissues tool kits.
Research data relating to studies which directly inform national policymaking should be considered for permanent preservation or deposit in an archive or repository: discussions may need to involve staff across the MRC, including experts in data management and research governance. In some cases the potential impact on policy may be a clear aim of the study, while in others the significance may only come to light later. It may therefore be necessary to consider the impact of the study at several stages during its life-cycle, particularly for studies which have long-term goals and which may be running for many years.
Retention, storage and archiving All primary/raw data and related materials retained, stored or archived should be recorded and held securely, in such a way as to allow them to be understood and used by others in future. This would include information about regulatory and ethical requirements relating to access and use.
Research data held electronically should be backed up regularly and duplicate copies held in a secure and accessible format where possible.
The digital continuity and future accessibility of electronic records and data should be considered. |
B.4 Research data (including images) should be recorded and retained. Retention periods should be informed by data management and quality assurance needs. Where primary/raw data are subsequently enhanced, original and enhanced data should be stored together. It is important to avoid the over-enhancement or over-interpretation of data and images.
B.5 It is essential to manage confidential identifiable data appropriately, including data associated with tissue and biological samples. A number of tool kits have been developed to provide a guide to the regulatory environment, including the MRC Data and Tissues tool kit which outlines the requirements for use of confidential identifiable data and human tissue samples in healthcare research. The tool kit route maps are aimed at guiding researchers through the planning and approvals process and provide practical help on legislative and good practice requirements as well as a summary of MRC policies in this area. Publication of any research data, including in Masters/Doctoral theses or in an accessible data repository, does not negate the need to retain primary/raw data.
B.6 Local procedures (for example, Standard Operating Procedures, protocols, etc) for all routine methods to be replicated across a study, together with associated risk assessments, should be documented systematically, in plain English and ideally in a standard format to ensure clarity, consistency and accuracy. Where there is more than one approved technique for any given procedure within the organisation, clear records should be kept on which were used. Where procedures change, they should be version controlled and the current version should be available and readily accessible to all staff, students and visiting workers.
B.7 Protocols for the use, calibration and maintenance of equipment, together with associated risk assessments, must be clearly documented to ensure optimal performance and research data quality. Where protocols change, they should be version controlled and the current version should be available and readily accessible. Instructions for the safe shutdown of equipment in case of emergency should be readily accessible. Such quality assurance measures as outlined in B.6 and B.7 help demonstrate the robustness and validity of research data.
B.8 To maximise public benefit, the MRC supports open access to the published outputs of research in a timely manner as well as initiatives that aim to extend access to research data and resources generated in the course of MRC-funded work. These may include the preservation and sharing of datasets and other relevant materials in line with the MRC’s policy on research data sharing. Where it is possible to share data and materials, guidance must be provided to ensure they are used appropriately, with proper regard for issues relating to consent, confidentiality and in accordance with any relevant data security guidelines, including guidance on encryption and data management, and other relevant conditions.
B.9 There should be clarity on the ownership and custodianship of research data, samples and related material used or created in the course of the research. Agreements should be used to clarify responsibilities, arrangements for access to data and managing permissions, including sharing with collaborators or with researchers who move away from the original organisation. Agreements should be developed at an early stage and should take account of requirements imposed by the research funder in their terms and conditions. Agreements might outline the responsibilities of steering groups, research investigators and team leaders in relation to these issues. Where personal data and/or biological samples are involved, the terms of the consent must be taken into account. Model agreements, such as those developed by UKCRC working groups, should be considered where relevant.
B.10 Where an MRC unit is due to close, or where programmes end and programme leaders or principal investigators transfer or retire, arrangements must be made in advance to support the retention and management of samples and data. This may include the transfer of custodianship to another individual within the unit or to another organisation and arrangements should detail provisions for access and eventual destruction. Where relevant, the Data Controller and/or the Designated Individual of the new organisation should be notified.
Further guidance on MRC policy on research data sharing, preservation and management can be found on the MRC website.
