By Pat Zinno
Director of infrastructure Services and Support
Based in Morristown, New Jersey, Atlantic Health has as its motto "The Passion to Lead." Our anchor hospitals, Morristown Memorial and Overlook Hospital, in Summit, are at the forefront of medicine in a half-dozen specialties, including cardiovascular care, neuroscience, pediatrics, and cancer care, and many of our physicians are leaders in their respective fields.
In 2001, Atlantic Health made a decision to use technology in a strategic way to extend this tradition of excellence. We set out to create a powerful, resilient, and adaptable IT infrastructure-a foundation, really-on which to deliver advanced services that "put people first," starting with our physicians and patients.
Mobility that aids productivity
A hospital's employees are, in essence, a mobile workforce that's contained within a building or campus. They're constantly moving around, which hampers productivity and makes communication incredibly inefficient. To address this dual challenge, we put a wireless umbrella over the entire campus and are rolling out mobility services. Computers-on-wheels (COWs) put patient information wherever the provider is, enabling bedside charting and provider order entry, along with accurate, real-time capture of test results.
We also deployed a VoIP-based communication badge. Incoming calls and pages are routed to the badge so providers can be reached directly. Outgoing calls are voice- activated and leverage the hospital's directory so they require no dialing. The badge also allows the wearer to receive text messages, such as an alert that a lab test has been completed. We estimate this service will save staff members 30 to 60 minutes a day by reducing phone tag and the use of the "sneaker net" to chase down people and information.
Reducing disruption for patients and families
We leverage our infrastructure to enhance the patient experience. To lessen the disruption caused by a hospital stay, we provide patients and family members with free public wireless access. On our patient portal, they can create personalized "care communication" web pages to update friends and family on the patient's condition. The portal also allows patients to select a doctor, pre-register for a hospital visit, and sign up for classes and newsletters.
Investments with a tangible impact
Underlying these capabilities is an infrastructure that has been systematically enhanced over the past several years. In addition to deploying the wireless network, we undertook several closely entwined initiatives that included migrating data off server-based captive storage and onto a storage area network (SAN), deploying a Fibre Channel between our two main sites 20 miles apart, and strengthening our disaster recovery capabilities.
In turn we were able to:
- Improve the availability of data by deploying a SAN, which has a lower drive failure rate than captive storage.
- Reduce the recovery time for mission-critical data from 36 hours to four hours or less.
- Slash data backup windows from 20 hours to six by switching to disk-to-disk backups.
- Save $600,000 in storage costs by improving storage utilization.
Business units help classify applications and data
In all these activities, Atlantic Health's application teams-who are our primary interface to the business units-have been instrumental in helping us define our infrastructure priorities for storage, networking, and disaster recovery. For example, the application directors helped classify and prioritize 160 applications in use throughout the network. This process led to a three-tier scheme that governs how information is delivered to users and how it is protected from possible threats.
Mission-critical applications include patient registration, medical charting, emergency room systems, and e-mail. For these applications, data is accessed via Fibre Channel to ensure rapid retrieval by busy providers. There is full server and SAN redundancy between the primary and backup sites to enable a maximum recovery time of four hours.
Business-critical applications include the payroll system, intranet, and materials and financial management. Data is called up via iSCSI connections, which are perfectly adequate for business users' retrieval needs. These applications have a one-day recovery window. Virtualization allows us to use excess capacity on multiple servers at the backup site to facilitate recovery.
Business-important applications include IT creature comforts such as monitoring software and auditing servers. These have more static types of information and a one- week recovery window, and are restored from tape.
Layered on top of this basic structure, information lifecycle management (ILM) allows us to classify different types of information for purposes of automating the data lifecycle- using storage in the most efficient way possible while continually creating space for more current information.
Again, business unit participation is vital. Input from our Human Resources and Risk Management teams helped us define automated processes for storing, archiving, and permanently deleting e-mail. We expect to achieve significant storage savings from having implemented a similar scheme for radiological images, which are massive consumers of storage. Based on needs articulated by the radiology staff, images are initially saved to high-performance storage so they can be quickly retrieved. After 90 days, when providers' retrieval needs tend to lessen, the images remain available but are archived to less costly content-addressed storage until the retention period expires.
Why SAN is superior
Consolidating data onto the SAN has allowed us to reduce our storage acquisition costs on a regular basis. In 2005, 20% of our data was centralized on the SAN and 80% was on servers with captive storage. At first glance, this approach seems cheaper per terabyte; however, our SAN utilization rate was 98%, compared to 40% for captive storage.
By shifting a greater share of our data to the SAN-about 70% today-we have saved approximately $600,000 in disk costs to date. Serendipitously, this change enhanced backup and disaster recovery capabilities by enabling us to do disk-to-disk backups and real-time data replication and mirroring between our primary and backup data centers. We continue to pursue infrastructure efficiencies such as using de-duplication to recover storage space.
Building on a solid foundation
With these critical building blocks in place, we are rolling out a number of new capabilities. Key initiatives include giving patients secure access to their medical records, revamping cardiology imaging using the knowledge gained from our radiology PACS deployment, and implementing physician office electronic medical records and bar-coded administration of medications. All these projects demonstrate our unwavering commitment to quality care, patient safety, and continued healthcare leadership.











