The Bay of Plenty District Health Board is relishing a brand new hospital campus and buildings in Tauranga. However, those physical changes came with technical developments as well — in the form of a new, resilient network infrastructure that will help the board deliver improved services to staff and patients.
Before the changes, the board operated a standard star network configuration with branches hanging off a central core. There were no redundant links to pick up service in case of outages.
As the network grew it was simply added on to, says the board’s general manager of information management, Owen Wallace. There wasn’t much network design going on and it was essentially used for data.
In 2000 a small voice over internet protocol system was introduced and the board began to think about converging its voice and data services. Plans for the new Tauranga Hospital campus also drove questions about future IT infrastructure, Wallace says.
Across the entire health system, the development of electronic medical records was also posing challenges for infrastructure.
The existing network was a mix of Cat 3 to Cat 5e cables. The upgrade was to address two main areas: the redesign of the core through to the distribution layer, and; upgrading the structural cabling in buildings to Cat 6.
In addition, fibre would be put in the ground around the campus and the standard of fibre improved.
IBM had delivered services to BoPDHB before the changes, but the board went to market to select a provider for the new project. In the end, IBM retained its position and was engaged for the core network build.
The cabling and fibre part of the project fell to the building contractors and local firms managed by BoPDHB staff.
A wireless element was also present, in the form of wireless bridges between buildings.
The network also had to serve the digital radiology system and systems such as patient telemetry and monitoring, as well as building security. With that amount of responsibility, the network had to be highly resilient to provide these services 24x7.
“It became clear the IP network was central to the way the organisation ran,” Wallace says.
To deliver resilience, IBM designed and implemented a medical-grade, Cisco-based network consisting of two core switches in separate locations. Both were designed to carry the board’s full data, image and voice traffic load.
The project came together “remarkably quickly”, Wallace says, with a kick-off in February 2008 and a design workshop in March. Wallace puts that speed of implementation down to a longer planning process.
Wallace says when IBM entered the picture, new ideas were introduced and there was a “meeting of minds”.
One of the biggest challenges was the need to keep services running throughout the implmentation.
“We couldn’t afford outages. We had to build and then cut over,” he says. “We didn’t have the option of turning the old core off and putting the new one in.”
Work on the network also had to be coordinated with the physical campus rebuild.
Wallace says the benefits of the new $1 million network are too numerous to calculate. The board spends $250,000 a day on salaries alone, he says. As health becomes more dependent on IT, the cost of outages or the inability to deliver systems is measured in lost productivity and loss of quality.
The network also supports increased IT centralisation.
“That’s obviously a more efficient way of delivering IT infrastructure to a health sector that needs to be as efficient as it can be — because it’s not the wealthiest health system in the world.”
Providing quality information to staff at the right time will enhance their decision making and improve outcomes, he says.
The end of the project, he says, was a huge anticlimax. The resilience test involved failing over from one core switch to the other. When it happened, no one even noticed. Even VoIP phone calls went uninterrupted.
“I expected a ‘hold your breath’ moment,” Wallace says.
• 1Gbit/s to the desktop
• Network designed for complete fail-over – a self-healing and self-defending network
• Convergence for clinical and administrative data, as well as voice calls for more than the current 1400 extensions
• Network traffic management system prioritises critical data
• Network will support deployment of new clinical applications
• Ability to centralise IT activities to boost efficiency and to roll out new services and applications