The Brisbane Southside HealthConnect Trial: Preliminary Results

The Brisbane Southside HealthConnect Trial (BSHCT) uses diabetes as a basis for a trial to test the value, technical feasibility and sustainability of an electronic health record shared between the public and private sector. This trial focuses on testing the HealthConnect architecture and the openEHR record architecture as an approach to underpin the communication and storage of health record data. The analysis and design phase of the trial has been completed, and the software is currently under construction with deployment planned for late 2004. This paper presents some early learnings. Early results point to the need for stakeholder engagement; cost effective recruitment and registration processes based on an opt-in model; business processes for identifying and correcting errors in the record and source systems; event summaries should be able to be reviewed by clinicians and consumers prior to submission and must be medico-legally safe, complementary to referral processes and are extractable from front end systems; when accessing the record we found that providers prefer to view status information extracted from event summaries, such as current medications, problem lists, etc, rather than the content of event summaries; the care team proved to be a useful concept for controlling access to the record and for quickly finding who is involved in a consumer’s care; and consent models need to be simple and sufficient whilst not adversely impacting on day-to-day clinical workflow. Introduction: Many governments around the world have recognized the benefit of a national shared electronic health record (EHR) and are now investing in EHR initiatives. A widely accessible shared EHR enables integrated care between the public and private sectors and across the acute, primary and community care settings. A widely shared EHR can deliver large scale savings as it can: lower the risk of adverse events, facilitate communication and coordination, enable clinicians to spend less time searching for or re-obtaining information and test results, and provide better information for planning of future health service requirements. HealthConnect is a federal, state and territory supported initiative to establish a national health information network to allow, with the individual health consumer’s permission, personal health information to be safely exchanged and stored within in a national shared EHR system. Over the last two years, the HealthConnect Program Office has undertaken a wide range of research and development activities in preparation for establishing HealthConnect [HCx2003a, HCx2003b and HCc2003c]. One of the major activities of HealthConnect has been to establish a number of trial sites to help inform the national architecture. Trial sites include those within Tasmania, Northern Territory, New South Wales and Queensland. This paper describes preliminary results from one of the Queensland trials: the Brisbane Southside HealthConnect Trial (BSHCT). The trial is initially planned to be a small pilot, but may be expanded if successful. The trial aims to recruit about 1000 consumers who visit participating clinics in the Brisbane South area, and involves about 100 providers including GPs, endocrinologists, ophthalmologists, diabetes educators, dieticians and podiatrists. Organizational level support for the trial comes from the HealthConnect Program Office, General Practice Computing Group (GPCG), Queensland Health, Princess Alexander Hospital, Mater Public Hospital, Brisbane South Division of General Practice (and GPs associated with the trial) and Brisbane Inner South Division of General Practice (and GPs associated with the trial). The trial has just completed its analysis and design phase and the software is now under construction. It is expected that the software will be operational by late 2004 and the trial will run for 9 months. The core parts of the BSHCT System include a Health Record System (HRS) and a Registration System deployed at the Queensland Health data centre. The HRS, which stores each consumer’s health record, supports key components of the HealthConnect architecture, including event summaries, EHR lists, views, reports and notifications [HCx2003a]. The Registration System is used by trial officers to manage the registration and authentication details of trial participants. Users with existing front end systems (such as Medical Director or IBA Spectrum (Plexus)) will be able to access the HRS using a combination of their local front end system and a Web browser. When a provider is viewing a consumer’s record on the local system they will be able to use a single click to view the equivalent record on the HRS. This works by the local system invoking a web browser to open the appropriate record on the HRS. Using a single click, provider will also be able create an event summary for secure upload into the HRS using information collected in the local system during the current session. Prior to upload the event summary can be edited by the provider. Other providers which do not have a local front end system, such as allied health users and ophthalmologists, will be able to view the consumer’s record using a web browser and to enter event summaries directly through Web based forms. In the trial, pathology data will not be collected directly, but instead it will be uploaded into the HRS as an event summary by the ordering GP deciding to forward the result. Consumers will be able to view their own health record via the Web. In the trial, all users will be authenticated using a user ID and pass phrase combination. As providers will be able to access multiple health records, they will also be required to authenticate with a HeSA location PKI key. Consumers can control which locations can access their record by selecting which locations are part of their “care team”. The system will also use a role based access control (RBAC) mechanism to further constrain access to the record. Figure 1: openEHR Models One of the main areas of focus of the BSHCT is testing openEHR as a potential record architecture for HealthConnect. A record architecture specifies “set of principles governing the logical structure and behaviour of health care record systems to enable communication of the whole or part of a health care record.” [CEN99]. In some systems this is implicit and proprietary, and in other systems it is open and standardized. Examples of open record architectures include CEN 13606 [CEN99], HL7 CDA [Dolin01], Good European Health Record (GEHR) and openEHR [openEHR04]. openEHR provides a harmonized architecture for the CEN 13606 and HL7 CDA standards. A key innovation in the openEHR architecture is that it separates record keeping concerns from clinical data collection concerns using archetypes [Beale02, Heard03]. As illustrated in Figure 1, openEHR uses a reference Instances Models Reference Model Archetype Model