Data Storage / Data Management
Critical to care
A comprehensive approach to data management in a hospital environment.
By Jim Beagle, June 2013
Hospital data is some of the most critical data to manage, yet it’s unwieldy due to its complexity and volume. Data growth in hospitals has escalated in recent years, accelerated by the move to eliminate paper and convert patient records into electronic files. Government mandates, including meaningful use and HIPAA/HITECH, have increased the urgency of this conversion as organizations seek to enable more benefits in patient care from the use of electronic data. They also hope to gain cost-savings advantages from centralized management of data. In practice, however, the complexity of hospital data systems presents some unique challenges and constraints, which tend to undermine these benefits. This combination of factors has brought health IT organizations under increasing pressure to find better methods for data management that fit the unique requirements of healthcare environments.
The data found in the electronic healthcare record (EHR) today spans many types, including data from healthcare information systems (HIS), radiologic systems (RIS) such as picture archiving and communication system (PACS) imaging applications, structured and unstructured files, and scanned documents. Even back-office application data, such as email messages, are part of the patient record and must be managed as such. A better method for data management in hospitals, therefore, must be one that enables management of all types of hospital data.
The key requirements of hospital applications in data management
Each hospital application has unique requirements for data availability and performance workflow, which must be considered when selecting a data-management approach. In the past, many hospital applications – notably PACS applications – tended to operate on their own. This allowed departments to manage data for their particular application in isolation. However, the data generated by department-level applications has grown to the point that this approach is no longer practical. A recent trend is for healthcare IT to start to transition under central management what had previously been isolated, departmental applications.
The intention of centralizing management is in part to obtain cost advantages in storage. In addition, more mature data-management approaches are possible. However, finding an approach to data management that can unify storage and data while still supporting the unique requirement of each type of hospital application is a tough challenge for most teams.
The importance of ensuring that the data-management approach selected is appropriate for each of the hospital applications must not be underestimated. The productivity of clinicians depends on application performance and workflow. Also, clinicians cannot tolerate any loss of clinical data nor extended recovery periods, as they risk patient care. A sufficient approach to healthcare data management must offer methods for data protection – purpose-built to integrate with individual hospital applications, yet unified to operate in concert across a centralized storage environment.
Sustained data growth adds unsustainable cost and administrative burden in hospital environments
The move to digital information as a replacement for paper is responsible for the creation of terabytes of new data every year at healthcare organizations of all sizes. A typical community hospital in 2013, for example, has multiple terabytes of data under management. This figure is expected to increase exponentially from year to year.
All of this data not only costs more to store, it also takes more time to manage than traditional approaches allow. Traditional backup on its own is no longer possible to accomplish in available windows. Even when data can be protected, recovery times take too long to be tolerated and data retention is hard or impossible to manage.
Lacking better alternatives, many health IT organizations have resorted to short-term tactical placement of more storage systems – to at least get the data stored – and have delayed strategic decisions on better approaches to data management. However, this short-term approach is not sustainable: At some point in the near future, the increasing cost of storage will exceed available IT budgets. It is imperative that healthcare IT teams begin strategic planning in the near future in order to take control of healthcare data in their environments.
Healthcare data growth, in itself, is no real surprise. So why is the challenge of managing these large volumes of healthcare data suddenly coming to light? There are two predominant trends that may offer an explanation:
- Hospitals’ move toward paperless systems: These projects are incentivized by government policy (such as ARRA and HITECH in the U.S.) and represent a major shift in most hospital environments. Both new and legacy patient records are being converted to digital form, which has ballooned data capture in healthcare environments.
- The massive increase in clinical images: An increasing number of clinical modalities are outputting a greater volume of medical images (based on the DICOM standard) in more frequent intervals and at higher fidelities. These images represent approximately 70 percent of a hospital’s total data footprint, according to a recent BridgeHead whitepaper.
Unquestionably, the conversion to digital information offers hospitals huge potential in the improved quality, efficiency and effectiveness of patient care delivery. However, for healthcare IT professionals, the situation is a challenging one, as it pertains to data and storage management. Therefore, new approaches are needed that can protect all healthcare data in a cost-efficient manner while ensuring the workflow and availability of healthcare applications.
Breaking down the data silos
Experts agree that it is imperative that healthcare IT teams find new ways to streamline access to patient data. With new requirements stemming from meeting federal initiatives tied to meaningful-use Stage 2 (MU2), interoperability and speedier access to patient data are critical for hospitals to achieve compliance. Jeanine Banks, GE general manager of global marketing for specialty solutions, recently noted that the industry is moving toward standardizing data for the sake of interoperability between different vendor EHR systems. Meaningful use and more efficient workflows in general are driving these technology initiatives that, at the same time, must promote HIPAA compliance. Unfortunately, the significant growth in healthcare data has made it difficult for some institutions to meet such requirements.
First, with massive data volumes comes the implied need for massive storage capacity. Storage silos found across hospital department systems have often meant poor utilization and higher pricing of individual storage systems. Often under management by clinical teams, these systems have suffered from the rapid growth of department data and are reaching critical breaking points. Increasingly, centralized healthcare IT teams are being asked to assume responsibility for these siloed applications, along with the data storage that supports them. While increasing the healthcare IT burden, the opportunity is that centralized teams with expertise in tiered storage management can help to reduce the cost burden on hospitals. Shared use of application data is also expected as a benefit for centralizing data management in these environments. Centralized data is also far easier to manage through its entire lifecycle, which includes adherence to long-term retention of patient data.
IT professionals are increasingly assuming responsibility for maintaining PACS applications that generate DICOM images. This image data cannot be protected using the traditional back-up methods, with which most IT professionals are familiar. The DICOM images are too large to be backed up in available windows. Also, the recovery methods enabled by traditional backup take too long to support the availability requirements of clinical radiologists. What’s required is a new, healthcare-integrated approach to clinical data protection for these applications.
Finding a new approach
By identifying and categorizing healthcare data as to whether it is dynamic (frequently accessed and/or changed) or static (where it will not change and is unlikely to be accessed after 90 days), as well as understanding the process by which dynamic data becomes static over time (and thereby has a limited useful lifespan – often for the duration of patient treatment), hospitals can employ different strategies to ensure the smooth, efficient and effective security, protection and ongoing management of their data. Hospitals can utilize the strengths of archiving, in conjunction with traditional back-up techniques, to assist in eliminating storage silos, optimizing storage assets, enabling data interoperability, ensuring full data protection and providing a rapid return on investment – all in a bid to put healthcare IT professionals back in control of their data and storage.
With the elimination of departmental storage silos, IT can begin the process of optimizing the storage assets of the entire organization. Economies of scale take effect as expensive tier-1 storage can be consolidated and leveraged by all of the applications that require it. Under the central management of IT, subsequent storage tiers can be deployed in a manner that maximizes their effectiveness by matching the value of the data to the underlying performance characteristics and availability of the storage on which it resides. With the aid of a central storage-management framework that cuts across applications and vendors, IT can repurpose – rather than eliminate – aging storage assets, thus freeing departmental budgets for other IT initiatives. When the time does come to retire existing storage assets and migrate data to new platforms, IT can perform these migrations seamlessly, without impact to end-users or applications.
In healthcare, consolidating the physical management of storage is only part of the benefit. Along with storage silos come data silos, which often hinder interoperability of systems. Without someone driving a strategic vision for enterprise-wide content access and sharing, healthcare information remains locked in these departmental silos, thus inhibiting the organization from realizing the full value of this data. With a truly healthcare-aware archive in place, the CIO can now collaborate with peer department heads to facilitate enhanced data interoperability of systems. To do this effectively, the archiving solution must leverage the healthcare standards by which these systems can interact and fully exploit the benefits of shared data. These standards include:
- HL7 (Health Level 7) for the exchange, integration, sharing and retrieval of electronic health information;
- DICOM (Digital Imaging and Communications in Medicine) for the storage and transmission of medical images and medical imaging data; and
- XDS/XDS-I (Cross Enterprise Document Sharing /for Imaging) for the sharing of clinical documents, images, diagnostic reports and related data.
In addition, the archive should have the ability to index both metadata and content to make that data easily searchable by both applications and end-users. Institutions that invest in a progressive deployment of common platforms can easily share data among systems, and rid themselves of the complexities of entrusting their data to multiple vendors and hindering access. For example, a Massachusetts-based hospital first invested in a vendor platform that ensures image availability at all times. Now, in order to build on that deployment, the institution is currently incorporating operational and disaster-recovery protection to further safeguard those image files. The hospital’s vendor supports the use of XDS-I for access to DICOM images and XDS.b for access to images that are not in DICOM format. This puts the hospital and personnel, such as its radiology team, in an excellent position to meet the requirements of MU2 pertaining to image sharing with patients and providers.
By eliminating storage silos and consolidating expensive primary storage, tier-1 storage assets are no longer underutilized. Thus, healthcare organizations do not pay for expensive storage that sits idle, and they enjoy more flexibility to employ less-expensive storage where the data access profile or data value supports that decision. A sufficient approach to data management must maintain storage-agnostic flexibility with policy-based archival to ensure cost-effective use of tiered storage systems. Only this unique combination will close the divide between what clinicians want and what IT needs in order to provide an intelligent approach to healthcare data management.
About the Author
Jim Beagle is president and CEO, BridgeHead Software. For more on BridgeHead Software: www.rsleads.com/306ht-201