White Paper: Understanding the Vendor Neutral Archive, February 2013 - by Joseph L. Marion

Executive summary

A changing healthcare legislative environment dictates that IT management be pro-active and creative in finding ways to improve diagnostic information accessibility and reduce costs. Vendor Neutral Archive (VNA) technology can be an effective tool for meeting data management and accessibility objectives. This white paper examines factors that can affect a VNA deployment as well as considerations for a strategic deployment of a VNA.

Having a clear understanding of what the VNA is, and how it best fits into a particular healthcare environment is key to a successful acquisition and implementation of the correct technology

Historically, data storage and viewing has been the responsibility of individual systems such as a Picture Archive and Communications System (PACS). Such devices were restrictive in that data management and access were dependent on system functionality. Some systems might be proprietary, complicating system change out and forcing expensive migrations.

Radiology has been the most-frequent service to employ PACS technology, but more recently other services such as cardiology have adopted the technology. Usually this has been through separate systems due to user preferences, creating multiple silos of data.

As healthcare legislation expands to encompass imaging, VNA devices will need to address more than just the DICOM (Digital Imaging and Communications) format, including other document and unstructured data formats. An added advantage of centrally handling this data is a greater opportunity to apply life-cycle management rules and to address data integrity, disaster recovery and business continuity objectives. The advent of EMR’s (Electronic Medical Records) means greater opportunity for data interoperability through a VNA, and simplifies accessibility by minimizing confusing viewing options. A VNA can further enhance a facility’s ability to participate in Health Information Exchange (HIE) and ACO (Accountable Care Organisation) initiatives by improving diagnostic data access.

VNA implementations can span from simple PACS archive replacement, to a consolidated multi-departmental archive, to enterprise and regional repositories. Requirements will vary depending on the objective, and it is important to fully understand a facility’s requirements before considering alternatives.

BridgeHead is comparatively new to the VNA application, but can capitalise on its extensive archival and EMR interoperability experience to provide VNA solutions that bridge both clinical and IT requirements. BridgeHead’s knowledge of data integrity and protection, as well as experience with a broad array of structured and unstructured data can enhance VNA options.

A successful VNA implementation is one that encompasses a common understanding of requirements and provides a platform that can meet today’s, and grow with tomorrow’s, requirements.

Introduction

Imaging and associated data are pervasive throughout the healthcare infrastructure. Until fairly recently, data management has been mostly diversified and the responsibility of individual imaging and document systems. With greater imaging proliferation, and the initial inclusion of imaging in changing healthcare legislation (ARRA/MU), healthcare Information Technology (IT) services are beginning to pay greater attention to image and document management. Of particular interest is how images and other patient content interoperate with the rest of the IT infrastructure.

This White Paper attempts to provide some clarification and direction to the consideration of a VNA, and how to ensure an approach that offers longer-term as well as immediate benefits from the implementation

One technology often touted as a means of addressing better image content management is that of the Vendor Neutral Archive, or VNA. Unfortunately, like so many other acronyms, the VNA can be greatly misunderstood. Having a clear understanding of what the VNA is, and how it best fits into a particular healthcare environment is key to a successful acquisition and implementation of the correct technology.

In addition, it is important to consider other data formats found within the hospital environment: Alongside DICOM images are non-DICOM images and unstructured files. In fact, the amount of unstructured data produced and stored in healthcare environments as part of the essential patient record rivals and in some environments outpaces that of DICOM images. Yet strategies for managing this full set of data have been largely ignored in discussions about the VNA. In this White Paper, we will attempt to see beyond the current horizon of what a VNA can do now and describe the potential for even greater benefits when a VNA-style of approach is used to manage unstructured hospital data alongside images.

This White Paper attempts to provide some clarification and direction to the consideration of a VNA, and how to ensure an approach that offers longer-term as well as immediate benefits from the implementation. Additionally, it will explore expanding the role of a VNA to determine its use and value when implemented to manage unstructured hospital data alongside images.

Before the VNA

Putting the VNA in proper perspective can be aided by an understanding of how content management has been handled prior to the VNA. Managing images has historically been the task of Picture Archive and Communications Systems, or PACS. These systems have grown up largely as responses to individual departmental service needs, and predominantly radiology services. A complete PACS managed the acquisition of images from imaging devices such as a Computed Tomography (CT) scanner, the workflow processes associated with storing and accessing the images for diagnostic interpretation, and the long-term retention of images for future comparison. In the early days of the technology, computer costs were relatively high and consolidating functionality on a singular platform was cost effective. Typically, the entire PACS was acquired from a single vendor, and hardware was maintained as part of a service maintenance contract. Vendors preferred this arrangement as it limited the number of hardware configurations they needed to support, and made interoperability among components simpler.

Architecture-wise, early PACS relied on spinning media for short-term image cache only, due to the relatively high cost. Post interpretation, most studies were stored long-term on less expensive media such as magnetic tape or optical disk. Workflow processes managed retrieval of prior studies upon demand so they were available for interpretation.

Accessibility to display images was oftentimes dependent on the PACS, both for diagnostic interpretation as well as clinical review. Review stations located in clinical areas such as the ICU or emergency department were all connected to the PACS. If the PACS went down, all ability to access and view images was lost.

As IT services consider their options, remote and cloud-based services may be part of the mix. Such services can address both capital acquisition limitations, as well as serve as bridge services to span capital appropriations cycles when additional capacity is required

Despite the evolution of an industry standard format known as the Digital Image and Communications (DICOM) standard, many PACS managed images internally in a proprietary format. That meant that they were only accessible via the PACS. Unless one understood the internal format, no 3rd Party device could access the images independent of the PACS.

This made update or replacement of the PACS an expensive proposition as the data needed to be “converted” or “migrated” from one format to another. A further complication to early PACS that might not have had an interface to a Radiology Information System, or RIS, was that study identification might have been by manual entry, further complicating subsequent matching of the study with a uniform patient identifier such as that provided by a Hospital Information System (HIS) or Electronic Medical Record (EMR) system. Migration of old records can be a slow and expensive process. It may occur either prior to the go-live of the new PACS, or over time by maintaining both systems during the transition – all adding to the cost.

Since most PACS were self-contained and did not have extensive interaction with the rest of the IT environment, basic IT practices such as disaster recovery and business continuity might be foreign to departmental practices, and therefore not high priorities. In addition, simple database backups were typically managed by departmental staff, and may not have ascribed to the same rigid policies of IT. Any system failure or disaster could result in extensive downtime and slow recovery processes. Hurricane Katrina was an eye opener for many imaging services that did not have adequate backup policies!

As time progressed, PACS have become better at interoperability, thanks to emerging standards such as DICOM, HL7, and IHE (Integrating the Healthcare Enterprise). Unfortunately, not all image content conforms to the DICOM standard. For example, some gastroenterology studies have associated sound that is not part of the standard but important to the images. Scanned or photographic images may already adhere to photographic standards such as TIF or BMP formats. Therefore, solutions need to address both DICOM as well as non-DICOM content.

PACS have largely remained departmental solutions. Hence, the advent of additional service areas such as cardiology, or ancillary imaging services typically were implemented as separate systems with their own acquisition, workflow, and image management applications. Departmental vendor preferences might also be a factor in the proliferation of systems, often exacerbated by the proprietary nature of vendor solutions.

Similarly, associated documents have been managed separately, either through commercial document management solutions or through home-grown applications. As with multiple image data silos, documents become just another data silo, albeit one that has been classically managed by IT.

Over the past several years, healthcare reform has accelerated, driven primarily by the ARRA/MU legislation. Stage 2 requirements of ARRA/MU have begun to address the inclusion of imaging, emphasizing image access via the EMR. The diversified nature of imaging creates challenges in terms of the number of locations to access images. While centralized, document management systems typically have no relationship or association with potentially-related image content, and represent another interface to an EMR. Having a single patient-centric solution would greatly enhance the ability to meet image accessibility requirements in ARRA/MU.

Why a VNA?

Given a background of factors leading up to a VNA, what might be the reasons a facility would be considering the acquisition of one? Part of the rationale behind this White Paper is the complexity of multiple scenarios for a VNA implementation, and the need for a proper understanding of requirements to assure that an implementation meets a facility’s needs.

Data Migration Mitigation

Implementing a VNA can be beneficial to an organisation, but it is crucial that there be a good understanding of organisational strategies and objectives to insure that all imaging service needs are taken into account to avoid replication of effort and redundant infrastructure

As indicated above, facilities struggle with the time and expense associated with data migration when a PACS is updated or changed. One rationale for a VNA is to create a data repository that does not have to change when an individual PACS is changed. Creating an environment that can accommodate multiple PACS can also avoid addressing the issue multiple times across multiple service areas. Case in point: a large academic centre in the New York City area was contemplating updates to both radiology and cardiology PACS. Initially, it was considering vendor-proposed migration costs for both service areas, as the PACS were from multiple vendors. Subsequent consideration shifted to a strategy to lead with a VNA implementation that would enable the sites to pro-actively move images from the existing PACS to the VNA prior to a change, for both radiology and cardiology services. Assuming there is sufficient time between implementation of the VNA and replacement of the PACS, the migration cost would be at best eliminated, and at worst minimised, based on the number of studies moved to the VNA. Such a strategy would greatly reduce the effort and expense to launch a fully-functioning replacement PACS.

Over time storage media technology can improve. Having the ability to easily migrate information from one media form to another without impacting end-user applications can be more timely and cost effective. Typically, VNA applications encompass such capability to migrate media in the background without impacting end user application access.

Economies of Scale

As previously stated, early PACS encompassed complete turnkey hardware and software solutions. One factor was the company’s ability to minimize the number of configurations to validate, thereby managing their cost. With the advent of the DICOM standard and the proliferation of storage media options, it has become substantially easier to interface multiple storage options. Storage media options have also expanded to include more options including tiered and mixed media and cloud options. Consequently, it may be simpler to capitalize on a facility’s existing data storage infrastructure for image management, rather than proliferating different solutions with each new PACS.

By relieving PACS of the long-term image archive role, facilities can make better use of their IT infrastructure investment. For example, a facility may budget additional storage capacity in a specific budget cycle. The capacity can then be optimized around the devices supported much more easily than if storage has to be budgeted and secured for each system separately. The net result is a more cost-effective management of a facility’s capital resources.

By uncoupling the long-term archive, facilities may have more flexibility in media alternatives that are not dependent on the PACS vendor’s approved devices. That way facilities can look to the open market to find storage solutions that provide the best cost-benefit for their needs.

Besides capital costs, facilities may have long-term support agreements with IT infrastructure vendors that can be leveraged at a substantially lower cost than what PACS vendors might offer, thereby increasing the savings potential.

Remote Managed Services and Cloud Hosting Options

As IT services consider their options, remote and cloud-based services may be part of the mix. Such services can address both capital acquisition limitations, as well as serve as bridge services to span capital appropriations cycles when additional capacity is required. There are multiple vendor options that offer both approaches for a VNA application. Facilities may want to consider a “try before you buy” approach to data storage, or consider remote/cloud services as part of a multi-tiered approach where off-site storage makes more sense for older, less frequently accessed data. Remote or cloud-based services can also be an effective means of “buffering” external data from internal data such as in the case of Outreach services where it would be inappropriate to intermingle external with internal studies.

Accommodate Multiple Data Formats

With the advent of EMR implementations, demand is increasing for expedient access to all of a patient’s information (images, proof of identity, insurance documentation, correspondence, etc.). One approach is to capitalise on a common storage infrastructure by partitioning it for individual system use. Another approach might be to create a common patient-centric storage infrastructure that can handle multiple data formats. For imaging, the DICOM standard has expanded to address many imaging devices, but as of yet not all devices. Therefore, it may be important to accommodate multiple data formats. This may allow integration between diagnostic images and documents within the same infrastructure, and accommodate imaging areas that are not yet addressed by the DICOM standard. A common approach has been to “wrap” non-DICOM images into a DICOM format by populating the DICOM header information and transforming the image content into DICOM. Unfortunately, this involves time and resource consuming transformations that may impact them overall efficiency of the system, may increase the storage size, and may be suspect in terms of concerns with image tampering.

BridgeHead is comparatively new to the VNA application, but can capitalise on its extensive archival and EMR interoperability experience to provide VNA solutions that bridge both clinical and IT requirements

Consider the plight of a large west coast medical centre that is in the process of implementing an EMR. The objective was to eliminate the patient’s paper chart. During the course of planning for implementation, the site realised that many diagnostic and surgical procedures today are captured, printed to hard copy, and stored in the patient’s paper chart. When the paper chart goes away, how will these images be accessed? Medical records indicated they can handle digitisation of documents, but not images, while existing PACS could scan them as secondary capture images. If managed within the PACS, the key issue will be how to associate them with the rest of the patient folder. If documents, associated images, and other unstructured content could be managed in the same storage solution, it might simplify access via the EMR. For example, consider an instance where there is email correspondence relating to diagnostic results. Would it not make sense to be able to have quick reference to this information along with the images and reports to provide a more complete diagnostic result? Accommodating multiple data formats within a VNA architecture may very well provide a means for doing so.

The advent of the IHE (Integrating the Healthcare Enterprise) profile for cross document sharing (XDS) has been extended to include imaging (XDS-i). The use of XDS-i in a VNA might be an approach to managing images as part of a larger image management solution.

Life Cycle Management

With the rapid increase in the amount of image information created throughout the enterprise, management of the information becomes significant, particularly when one considers healthcare legislation such as the HIPAA (Health Information Privacy and Accountability Act). HIPAA provides for the protection of Patient Health Information (PHI) in the healthcare enterprise. Having patient information managed in multiple devices further complicates information management.

Consider the aspect of image life cycle management. Almost all image information is regulated by retention requirements, typically five to seven years. At the end of that period, the images may be discarded if no further patient activity has occurred. If image information is spread across multiple devices, managing the deletion of that information can be a daunting process.

On the other hand, if the images are managed centrally such as in a VNA, a common retention algorithm can be applied to delete the images at the appropriate time. Occasionally rules change, and again having the data in a single location can make rule changes simpler to enforce. Of course this requires that PACS and other related systems can accommodate a centrally managed archive in terms of proper handshaking to support status updates when a study is deleted.

Data Integrity

Managing large amounts of data involves multiple levels of an archive, including the physical, abstraction, file and content layers. Most PACS and some VNA’s only address the upper layer (content) and do not address how data integrity is maintained. Most IT organizations understand that there is such a thing as silent data corruption, whereby the data is corrupted when written to the media. For example, a hard disk drive firmware might become corrupted and impact the way data is written to the drive. Consequently, even though the content layer may assume the data is in the archive, in reality it may be corrupted and irretrievable. VNA solutions need to take this into account and have the ability to assure the data integrity of content as well.

Disaster Recovery/Business Continuity

Most PACS have not adequately addressed the questions of disaster recovery and business continuity. Standard IT terminology such as RO (Recovery Point Objective) and RTO (Recovery Time Objective) may be foreign to PACS vendors, as evidenced by RFP (Request for Proposal) responses experienced by the author. A VNA can enhance the ability to address disaster recovery by providing a singular source to restore (versus a number of separate PACS). Additionally, a VNA that can be accessed independent of a PACS can provide business continuity during a PACS outage by providing an alternate path to image access by means of a separate viewing technology.

A VNA can provide a valuable buffer to a PACS in that the PACS may continue to operate even while the VNA is being recovered – something that may not be possible by the PACS alone. If the VNA is part of the IT infrastructure, RPO and RTO objectives may be more aggressive and more effectively achieved, as opposed to individual PACS. All are important factors in achieving data accessibility and quality of care objectives.

EMR Accessibility

A changing healthcare legislative environment dictates that IT management be pro-active and creative in finding ways to improve diagnostic information accessibility and reduce costs

One of the key advantages to an EMR is the ability to access all diagnostic information on a patient from a singular source. The objective of an EMR is to provide simple access to the complete patient record. If image and documents are managed by multiple devices, accessing and displaying content might require multiple image viewer technologies. Physician satisfaction and acceptance of an EMR is predicated on ease of use. If a physician were required to learn multiple viewer technologies, it could adversely impact acceptance. EMR integration of a common image viewer in conjunction with a VNA can simplify EMR image access and potentially accommodate multiple image formats.

The latest viewers encompass zero footprint technology that enables image display on any device that supports a browser, including tablets and smart phones, thereby extending the accessibility to clinical information.

Health Information Exchanges

Healthcare legislation has stated objective to lower the cost of healthcare. In areas where patients may have competitive healthcare provider choices, and with the advent of Accountable Care Organiastions (ACO), providing access to patient diagnostic information across healthcare enterprises is becoming important. Participation in these so-called Health Information Exchanges (HIE) can be enhanced if patient information at any given facility is readily accessible. VNA’s can help on this front by providing a central access point for image and document content, simplifying a facility’s ability to participate.

There are classically two models to an HIE: the federated versus the centralised model. In the case of the centralized model, patient data may be stored in a regional archive that is accessible to multiple participating entities. In the case of a federated model, the same result can be achieved by providing a common view to multiple storage devices, meaning a participating facility might maintain a VNA locally, but enable other participants access via a central registry. In either event, a VNA can play a key role in enabling a facility to achieve integration across multiple entities.

VNA Strategies

As discussed above, there are multiple reasons for consideration of a VNA. Having a clear understanding of those reasons is critical to a successful VNA implementation. The following examples illustrate potential levels of a VNA implementation. Which one is right for a particular organisation will be dependent on the priority of departmental and IT initiatives.

Level 1: PACS Replacement Archive

Replacing a PACS archive is usually a departmental service area focus. The key objective is to augment or replace an existing PACS archive. Since most PACS rely on the DICOM standard for imaging, a departmental solution most likely will be DICOM based in terms of image content. One objective in implementing a VNA might be to replace a PACS archive that is in a proprietary format, and thereby minimise subsequent migration, as well as enable third-party devices the ability to directly interface the device.

Another motive for implementing a level 1 VNA might be to provide greater flexibility in adding other devices such as an advanced visualisation application. The ability to pull images directly from the VNA independent of the PACS can be valuable – especially if the PACS archive is in a proprietary format. The advent of a VNA also affords the option to consider alternative workflow solutions that can directly interact with the VNA, and update the departmental workflow without impacting the existing PACS. For example, this might be an effective means for implementation of a decision support solution.

Departmental VNA solutions might take advantage of off-site and/or cloud applications for disaster recovery, and may provide some amount of business continuity in the event of a disaster. These capabilities may be easier to add as part of a VNA than by updating the PACS.

Level 2: Shared Archival Resources

When multiple departmental services are involved, the VNA objective may be a shared archival resource. Typically the motivation comes from IT services that are looking to consolidate infrastructure to use common archive resources. A shared VNA resource offers advantages over and above just sharing or partitioning common archive infrastructure, most notably the ability to apply common management rules and media migration practices.

Typically such initiatives are built around DICOM based services, so they can be DICOM based. Another motivation for a shared resource VNA is to minimize migration effort. Economies of scale can also be a strong motivator in terms of consolidation of service maintenance agreements, and leveraging vendor capital acquisitions. As with departmental solutions, they may encompass some degree of remote/cloud-based storage if these are part of the enterprise infrastructure.

Level 3: Enterprise Image Management

Enterprise initiatives such as an EMR are usually the impetus to consider an enterprise-scale VNA solution. The key differentiation is the interoperability with the EMR and the need to potentially address DICOM, other image formats, and documents. Timing and strategy may also dictate consideration of shared capability with document management applications to achieve greater interoperability and economies of scale.

Usually such initiatives will encompass a common image viewer that interacts with the EMR and provides a unified patient-centric view of a patient’s diagnostic information. A common image viewer enhances user acceptance as it simplified how diagnostic information is viewed. As many “enterprises” involve more than one facility, most likely an enterprise VNA may involve remote/cloud services.

Level 4: Regional Repository Participation

Community-wide initiatives as part of Health Information Exchanges (HIE) or Accountable Care Organisations (ACO) may eventually address imaging. Preparing for an HIE or ACO may be enhanced by implementation of a VNA at individual sites that can become part of a federated imaging solution. It may be advantageous to interoperability to have a single image archive at each participating location versus interfacing multiple devices at each site.

Organisationally, an HIE may be structured as a separate entity from individual facilities. Simplifying how the HIE interacts with individual facilities may be important to the overall success of the HIE. By its very nature, an HIE and potentially an ACO may require off-site or cloud-based applications to provide a means for connecting all facilities.

VNA Architecture

VNA architecture will vary by vendor, but fundamentally the VNA portion is usually “middleware” that represents a bundle of services as illustrated in Figure 1. Most VNA’s encompass a number of services that address interfacing with various image sources such as modalities, PACS, or other applications. Once acquired the VNA includes a number of services for managing the workflow of image archive, including lifecycle management, replication, dynamic tag morphing (ability to change DICOM metadata content), etc. that constitute how the VNA handles the image data. Lastly, the VNA provides services for how the data is handled within the image storage infrastructure, such as which media the data is stored on, migration of data from one media to another, interaction with a cloud service, etc.

Some vendors package both hardware and software in a turnkey product, while other vendors remain hardware agnostic and provide only the “middleware” application. Choosing an option depends on individual IT circumstances including vendor relationships and hardware policies.

BridgeHead and the VNA

While comparatively new to the VNA arena, BridgeHead Software has a long-established track record in healthcare data management that affords it many advantages in terms of VNA interoperability. In a recent White Paper, BridgeHead lays out its perspective to some of the problems with the traditional VNA (White Paper: VNA Does Not Equal Image Availability: What You Need to Know - click here). The paper identifies many of the points made in this discussion, in terms of the need for solutions that address the full scope of data management and integrity.

Common points to which the author subscribes include:

• Focusing exclusively on the content layer of archiving omits archive functionality that ism required to ensure image availability
• Correctly asserts that a complete VNA strategy spans all layers of an archive, including Physical, Abstraction, File, and Content
• Is correct to point out the risks of silent data corruption, which can destroy the integrity of stored images
• Makes the case that a complete definition of a VNA must include support for all content types, not just DICOM
• Is correct to point out that the definition of a VNA must support both IT and clinical needs (e.g. relevant priors workflow vs. IT products that might not be designed specifically form clinical applications, plus support for multiple departments)
• Understands that data must be stored for long periods of time that may exceed the life of the storage media/device (i.e. storage device agnostic)
• BridgeHead’s approach effectively bridges the gap between what clinicians want and what IT needs

Many of these points apply to all levels of a VNA, particularly those having to do with data integrity. For higher-level VNA applications, BridgeHead is uniquely qualified based on its ability to handle greater content, and its experience with archival management. BridgeHead also has experience with EMR interoperability that is important to a VNA integration.

Conclusion

Implementing a VNA can be beneficial to an organisation, but it is crucial that there be a good understanding of organisational strategies and objectives to insure that all imaging service needs are taken into account to avoid replication of effort and redundant infrastructure. A well planned implementation can lower operational and maintenance costs, minimise migration time and expense, and avoid the complications of proprietary data storage. At the same time a VNA can be instrumental in the improvement of clinical data accessibility which may be a necessary part of an ARRA/MU strategy. A VNA strategy can also be a gateway to expanding service area capabilities without the potential need to replace legacy PACS.

A successful implementation is one that starts with a common understanding of objectives and assures that IT and all service areas are represented. Proper planning today can avoid future conflicts and assure that a facility meets clinical and financial goals in a changing healthcare environment.