This project relates to the research collaboration between Edinburgh Napier University, CipherLab, Chelsea and Westminster Hospital, GS1 UK, Imperial College, and Kodit, and is funded through a research project with the TSB (Technology Strategy Board). It focuses on the development of a highly secure e-Health Cloud (DACAR), which aims to improve the clinical environment with a completely scalable infrastructure, and which integrates identity, role and access rights to every part of the capture, storage and consumption of health care data. This work thus aims to produce a scalable cloud-based infrastructure that could be used in many different application domains.
A key challenge in health care is to use captured patient data in multiple forms of context, while still maintaining strict access rights, where access to the core data might be allowed for one application, and not to another. This might including using patient location data to determine patient pathways, and also in using the same data to determine the utilization of various resources within a hospital. Unfortunately, though, the original data is often lost when it is aggregated with other forms of information. Thus, in order to provide information in multiple contexts and to reduce data loss, the captured data within DACAR is stored in its original form, and marked-up with the required context of the patient data capture environment, so that the resulting information can be view in many different ways. Along with this, the rights of access to the data depends is based on a well-defined overarching security policy, and is managed through service interfaces, which has a strong definition of checking identity and role, and assigning rights based on these. Figure 1 provides an overview of the key focus of the work which aims to overcome the static and localized nature of patient records, and aims to produce patient care records which are dynamic, and which can be made up of any clinical data, but which is carefully controlled by a security policy.
Figure 1: http://buchananweb.co.uk/dacar01.png
Figure 2 shows an outline of the key elements of the infrastructure: Data Capture; Clou Storage; and Clinical Services. Captured data is securely stored in patient data buckets within a Cloud, using interlocked patient and service encryption keys. In order to control access to data, DACAR then does not allow direct access to the patient data bucket, apart from through exposed well-managed service interfaces, which check for identify, roles and rights. For example a patient may have clinical data for heart rate and blood pressure measures, but access to these can then be limited to an aggregated risk assessment score, of which rights are assigned from the patient (the owner) to their trust circle (typically their clinician). The system is thus patient-centric, where the rights of access to data is defined by the patient, and thus access rights can be removed at any given time.
The DACAR architecture consists of a series of SPoC (Single Point of Contact) which manage a given domain (such as a single health care organisation), and is used to manage requests by users within their domain, and between domains. A single security policy, which defines roles and rights, is then used to create a single and thus consistent security policy across the whole infrastructure. The SPoC thus checks identities and roles and grants access to services based on this. A key factor of the Cloud is the integration of an ontology for the complete infrastructure which defines roles, identifies, and rights for capture, storage and consumption. The access to data is then managed through well-defined and managed service interfaces, which integrate role and identify.
Figure 2: http://buchananweb.co.uk/dacar02.png