Implementing EHR applications: a Canadian experience

Recently, I participated in the implementation of Cerner’s Millennium clinical information system at a major academic teaching hospital in Canada; this software is now being implemented in the Southern and London clusters of the NPfIT. I worked as an application specialist during the hospital’s implementation and/or maintenance of the admission/discharge/transfer (ADT) component of UK PAS, results reporting, order communications, clinical documentation and pharmacy. The hospital took a traditional approach to implementation: all design, building and testing was done by the hospital’s inhouse informatics department (with consulting support from the supplier). The informatics team included clinicians from a wide range of backgrounds and staff throughout the hospital were involved in the design process — from participating in associated committees to completing applied testing scenarios.

Canada’s 10 provinces and three territories are constitutionally responsible for the administration and delivery of healthcare services, while the federal government employs financial incentives to set national standards, particularly those enshrined in the five principles of the Canada Health Act: public administration, comprehensiveness, universality, portability and accessibility.

The federal government lacks the authority to enforce the standardised procurement and implementation of systems at a national level. In 2001, it established an independent, not-for-profit corporation (led by senior healthcare representatives from each province), Canada Health Infoway, to develop a strategy to support and accelerate the development of electronic healthcare records (EHRs) on a pan-Canadian basis. Infoway has adopted a strategic investor role: it establishes the strategic direction for EHRs, invests in provincial and/or regional initiatives that are aligned with this direction and seeks to make better use of existing solutions where possible. Key investment programmes include registries, laboratory information systems and diagnostic imaging. Infoway aims to have EHRs in place for 50% of Canadians by 2009. A 2006 Commonwealth Fund study found that only 23% of primary care doctors in Canada use electronic medical records, compared with 89% of GPs in the UK, so GP systems cannot be a basis for Canadian EHR development as envisaged, for example, in Wales and Scotland.

Despite Infoway’s strategic investor role, its influence on local healthcare providers varies. Acute-care providers, particularly those that are not integrated within a broader regional delivery structure, remain relatively independent as to what clinical information systems they procure and implement, and a range of suppliers are represented across the country, including Meditech, Eclipsys, Cerner, Misys, McKesson, and EPIC.

Comparison of approaches between England and Canada

While both countries have identified value in increasing the use of ICT in their respective healthcare systems, England and Canada have adopted different approaches to these investments. Canada focuses on the development of a patient-centred pan-Canadian electronic healthcare record, while England has conceptualised it as a service rather than a system.

Department of Health procurements for its National Programme were based on a single standard set of requirements, the output-based specifications, whereas Canadian jurisdictions have led their own procurement efforts and investment decisions in keeping with their own strategic priorities, applying for Infoway’s financial support where they recognise alignment with the national direction.

In England, Care Records Service (CRS) functionality is being configured at a regional level, leaving limited scope for customisation to meet local needs and priorities (although this is expected to change with the new NPfIT Local Ownership Programme). Investment priorities between the two countries are different: England places a greater emphasis on systems to support primary care processes — eg, GP referral to secondary care through Choose and Book and electronic transmission of prescriptions, while in Canada there is a significant investment in ‘telehealth’.

Challenges faced at a Canadian hospital

The Canadian hospital at which I worked faced a varied set of implementation challenges, particularly related to the more complex clinical functionality.

Interfacing

Developing and testing interfaces between systems from different suppliers (eg, for results reporting and order communications if ancillary systems are procured from other suppliers) is challenging and time consuming, and may negatively affect the functionality available for use within the core software. England’s interfacing challenges could, perhaps, extend further, as field authorities seek a way of implementing interfaces between national and local systems. Infoway has not developed any national applications, but plans to use investments in patient registries to enable interoperability between disparate systems.

While complexity and scope may mitigate against a big-bang approach to implementing certain modules in a trust, piecemeal implementation also creates significant challenges, particularly where patients move from automated to non-automated wards. Rollout considerations also need to include the staffing requirements of supporting a full organisation-wide go-live of a system such as order communications, as well as what flexibility is available for other ICT projects.

Implementing clinical information systems is generally difficult for a number of reasons that include parallel governance structures (clinical and managerial); the limitations of available software; and the inherent challenge of automating elements of complex clinical processes in a co-ordinated and beneficial manner. Engagement challenges in England appear to have been exacerbated by the centralised nature of the Programme — fortunately, the process of engaging people in the field with central initiatives is not an issue in Canada, where healthcare organisations remain the drivers of their own initiatives and are fully accountable for their success.

Inputting site-specific data is one of the simplest tasks in implementing clinical information systems; the greater challenges lie in understanding how applications work, testing, finding problems, and working with suppliers to solve them. Reducing reliance on suppliers by creating inhouse capacity to deal with developments is invaluable — for example, investing in training ICT staff to write CCL code (the Cerner version for SQL) so that future customised reporting can be developed inhouse.

While the importance of change management should not be underestimated, it is also remarkably facilitated by presenting users with systems that meet their needs, rather than forcing them into a series of workarounds.

Conclusion

The general approach of the CRS component of NPfIT in its first four years — with its standardised application design and build, and generic functionality rollout timetable — differs significantly from the flexible and customised approach taken to implementing the same software in a Canadian teaching hospital environment. However, a single hospital can struggle to communicate the importance of its enhancement requests to the supplier, while the NPfIT benefits from its ability to take advantage of the size of the programme to drive system functionality improvements. Canada could learn from England’s bolder approach to eHealth, while England could learn from Canada’s approach to building on existing investments.

Universally, clinicians and patients benefit from the ability to access and share an integrated view of a patient’s medical information easily. The challenge for health informatics professionals is to help realise these benefits sensibly and affordably. Learning from other countries offers a broader perspective to working through these complex issues.