ABHI Position on Data Sharing
Basic principles and considerations
- International standards should be used to support a shared health dataspace, respecting local health system rules and legislation, to facilitate secondary use of aggregated data for research and the implementation of services.
- Data is reusable and cannot, therefore, be traded on a simple transactional basis.
- Raw data is, of itself, of limited value and the Intellectual Property should accrue to those that are creating value. We believe in competitive markets to drive innovation and economic growth, and actors involved in that should get fair reward relative to their inputs.
- Identifiable data should not be shared without the explicit consent of the individuals concerned.
- Anonymised aggregated datasets should be available to benefit patient care and system efficiency, and generate economic growth.
- There should be assumed consent for direct care, but patients should be made aware of any further possible use of their data.
- Specific consent should be gained for use of data outside of clinical research.
Beyond these basics, there are some specifics relating to ethics and consent, Intellectual Property and Data Access and Portability.
Ethics and consent: (to a large extent the issues here are no different to national issues on data sharing)
- What consent has the patient given?
- What is the data being used for: (care, commercial research, marketing?)
- How is the data being treated: (de-identification and aggregation?)
- What is the feedback to patients on subsequent uses?
Intellectual Property: (centres around fair share of value and business model)
- On what commercial basis is data being made available; (pay to access or share of value in output?)
- What services (e.g. data cleansing, curation, combining datasets) is each party offering?
- Who is creating the value?
Data Access: (is there a specific need to allow access to data and does ownership change?)
- What is the model: (access to data or use of data to answer specific questions?)
- Does data need to transfer to extract the value?
- How has the data been treated: (has it been de-identified and aggregated and is that process irreversible?)
Portability: (there is general principle that there is potential benefit in larger and appropriately aggregated data sets)
- How will data be exchanged and aggregated? (Combining data sets is rarely straightforward)
- Under which jurisdiction will shared data be controlled?
- What standards will be used? (Propriety standards should be avoided).