Healthcare storage virtualization is a flexible architecture that can help hospitals deal with the data explosion in a cost-effective manner.

Healthcare organizations are in the throes of a data explosion. Government incentives — such as the HITECH provision of the American Recovery and Reinvestment Act (ARRA), Canada Health Infoway and the European Institute for Health Records — are encouraging hospitals everywhere to adopt electronic health records (EHRs). At the same time, there is an upsurge in the creation of DICOM (digital imaging and communications in medicine) data. More and more hospitals are implementing new PACS (picture archiving and communication systems) to support not only radiology but also other clinical disciplines, such as mammography, cardiology, pathology, endoscopy and ophthalmology.

As a result, hospitals are finding it increasingly difficult and costly to manage all of this data; many are trying to plug the flow of digital information by buying more expensive disk space and parking their data on proprietary content-addressable storage (CAS) systems.

Even worse, they are not meeting their backup window requirements for disaster recovery. There is simply too much data and not enough time to replicate it all. These organizations may be meeting the federal requirements for maintaining digital patient information, but they are failing dismally in their most basic obligations to protect themselves — and patient care — against the event of a systems outage. And they are breaking their budgets, too.
Why is healthcare data management so hard? Is there a way to fix this problem?

To start, there are intrinsic differences between healthcare and traditional enterprise data. On a daily basis, hospitals deal with three different kinds of digital information: structured, unstructured and semi-structured data. The preponderance of unstructured and semi-structured data generated by PACS, and the speed at which that data multiplies, is unmatched by any other industry. By adopting a healthcare storage virtualization (HSV) architecture, hospitals can implement separate management policies for each content type and begin to get control of it.

HSV is a flexible data and storage management platform that decouples software applications from the allocation and management of the physical storage hardware on which the application data resides. By separating these applications from their respective storage devices, healthcare organizations have more choice, flexibility and control over the way data is accessed, protected and managed. The aim, ultimately, is to ensure optimal efficiency and cost effectiveness through better utilization of hardware resources and, more importantly, the easy, safe and reliable access, use and management of critical electronic patient data.

Because it is hardware agnostic, HSV architecture frees hospitals from the vendor “lock in” of proprietary CAS systems. It also allows hospitals to take advantage of their existing environments and only add new hardware as their budgets permit.
HSV architecture can also help resolve the thorniest issues of healthcare data and storage management. From data integration and the sharing of content between the various hospital information systems (HIS), PACS and document management systems, through to data migration and onto new hardware, HSV offers a comprehensive solution specifically designed for healthcare environments. It even helps shrink backup windows by reducing the amount of data to be replicated. Here's how it works:
First, the HSV management software identifies and characterizes the data. Then each data type is stored on the most appropriate hardware as best befits the needs of the healthcare organization.

Structured data, such as that generated by the HIS, is database driven and may reside in MUMPS, Cache, MAGIC, Oracle or SQL. RIS (radiology information system), LIS (laboratory information system), EHR and accounting systems typically fall into this category, as well. Structured data can comprise as much as 30 percent of a hospital's overall data. Since it is updated frequently, this data is allocated to primary disk storage, where it can be quickly accessed and replicated on a regular basis to meet the organization's disaster-recovery requirements.

Unstructured data is located in discrete files outside of databases; good examples are office productivity suites, such as word processing documents and spreadsheets. Often, multiple copies of each file are generated by multiple users. Unstructured data also may be audio and video files or photographs, such as a patient driver's license stored for ID purposes. Over time, these files can balloon into many terabytes (TBs) of data that can extend the data backup window beyond reason.
Semi-structured data is generated by PACS and ECM (enterprise content management) systems. Both maintain a database of information (structured data) that references often large quantities of discrete files (unstructured data). A PACS database may run on Oracle or SQL, and its size may be relatively small in relation to the many TBs of DICOM images that it references.

In hospitals, 70 to 80 percent of data is unstructured or semi-structured. After 14 months, the majority of this data will never be recalled or used again. Yet, by law, the hospital is required to maintain that data for the life of the patient or longer.
HSV recognizes these requirements and deals with large volumes of unstructured and semi-structured data accordingly. Consequently, this data is archived progressively onto lower-cost storage media (disk or tape), where it can be readily accessed if needed. The archive is automatically backed up for security purposes, but since the data will not change, it's removed from the ongoing backup process. By reducing backup volumes by as much as 80 percent, HSV transforms disaster recovery from a strategy to an ongoing, achievable process.

Finally, HSV helps hospitals go green. It moves large volumes of data off expensive disk storage, which needs constant cooling and backup, to archived storage hardware, which is less expensive and requires minimal maintenance. By getting a handle on the majority of immutable data stored by the hospital and moving it to archives, it also puts a damper on frequent disk storage purchases.
There is a way to effectively manage healthcare data; it's an architecture called healthcare storage virtualization — and it's ready to be implemented now.



Tony Cotterill is CEO, BridgeHead Software.
For more information on
BridgeHead Software solutions:



The Source for Healthcare Information Systems Solutions