Enterprise Archival is Not Well Understood
I am currently engaged with a number of clients that are in the midst of wrestling with imaging requirements in a number of service areas. In one particular facility, the radiology PACS is rapidly running out of long-term archive capacity and the physical storage is on an obsolete DVD Juke Box. Cardiology is just initiating a PACS, and of course, it is a totally separate vendor from radiology. In this case, IT is struggling with whether to implement updates separately, or consider an enterprise archive solution.
In the course of my consulting experiences, it is increasingly obvious to me that the industry in general is in its infancy in understanding enterprise image archival, and its value to the Electronic Health Record (EHR). I thought it might be helpful to share some viewpoints about enterprise archival, and perhaps raise the awareness level a bit.
There are multiple approaches to addressing image archival, including service area specific, virtualized enterprise archive, and middleware applications. In the case of service area specific archival, each service area solution manages its own archive. There may be operational advantages to this approach in terms of performance, particularly when the data is managed in a proprietary format. The downside is that multiple devices need to be managed, including disaster recovery approaches.
Many facilities have recognized the limitations to this approach and have worked to create an enterprise solution that accommodates multiple service areas in one physical storage solution. Each service area can maintain its own data if the virtual storage is partitioned. The storage media can range from simple storage devices like a SAN, to more complex approaches like Grid Mass Storage solutions. Such an approach addresses the limitations of a service area exclusive solution in terms of economies of scale and disaster recovery management, but it does not address a true enterprise view to the data as each PACS still manages its archive separately.
The alternative is to uncouple the individual PACS from the storage environment by introducing some form of “middleware” that can manage the archive device and service each PACS. In this case, all of the benefits of a centralized storage are achieved, with the addition of the added benefit of managing a central image repository that is accessible by enterprise applications such as an EMR.
Which approach is the right one for your facility? Among factors to consider:
§ How many service areas are currently managing digital images, and how many additional service areas are likely to do so in the near future?
§ At what point are you in implementing an EMR, and is imaging an important element in your EMR plan?
§ Are there reasons to consider accessibility to the information outside of the individual PACS applications?
§ Are there limitations to an independent archive, such as service area PACS vendor restrictions on 3rd party device validation or support contract limitations?
§ If changes are planned, how much data will need to be migrated from existing vendor solutions, and who will take ownership for migration?
§ How important is it to be able to independently manage the archival data, such as for patient demographic updates, or managing patient record retention rules?
Enterprise Archive Benefits
If any of the above considerations are important, an enterprise archive approach may be worth considering. Among the benefits to be derived:
§ Improved image storage economies of scale: By consolidating storage to a single approach, it may be possible to significantly reduce the cost of overall storage for both the physical media, as well as support costs. Additional savings may be achieved through elimination of older storage technology that may be obsolete and expensive to maintain.
§ A centralized archive can buffer applications from external users: Many facilities are interested in expanding outreach services as a means of growth. IT may have reservations about allowing external physician access to its applications, due to the risk of security breaches and malware intrusion. Also, how does one create an environment where an external facility can manage its data independently, but also allow access to referral studies done at a central medical center facility? The concept of an enterprise image archive can help by buffering individual applications from external users, yet provide a convenient access methodology to diagnostic results. It may also be easier to restrict accessibility to information without proper authorization.
§ Promotes Life Cycle Image Management: With service area specific solutions, or centralized solutions managed by service area specific PACS, purging or redefining purge rules can be complicated and inconsistent. An enterprise-scale 3rd party archive application that consolidates patient images from a number of service areas can simplify study purging by applying a consistent rule set to one instance of the patient information, and by enabling rule periodic changes to be made. The advent of digital imaging has changed classical image purging procedures, as many facilities that previously periodically purged film are handling digital images differently. They are either simply not purging studies, or old film-based studies are not retained because it was not clear there is a newer digital study which may or may not trigger retrieval of the film folder. A central repository makes it easier to check, and rule sets may be able to handle unique circumstances that might not be implemented for the service area archive. Of course, an enterprise approach is also dependent on the ability of the middleware archive management application to keep the individual service area applications appraised of purging activity and changes.
§ Migration simplification: An enterprise archive approach managed by a middleware application may be able to improve data migration. Over time, new storage media may become available that offers better performance at a lower cost. Taking advantage of a new media could be a daunting task if it has to be applied across multiple systems. Allowing a middleware application to manage the migration of data from one media to another can be done seamlessly in the background, without disrupting individual service area systems. Additionally, storing images in a centralized archive can significantly enhance the ability of any single service area’s decision to change a PACS vendor, as historical data may already be stored in a DICOM format that is accessible by the new PACS vendor, eliminating an expensive migration.
§ Ability to manage multiple data formats: In many imaging service areas like radiology and cardiology, the DICOM image standard is well implemented and supported, while other service areas such as Gastroenterology and Pathology have yet to be completely incorporated into the standard. Therefore, an enterprise archival solution that accommodates both DICOM as well as other image formats such as JPEG, TIF, etc. may have additional enterprise benefits in terms of patient information accessibility.
§ Simplified Disaster Recovery: HIPAA security regulations place greater emphasis on digital image management that can have severe consequences for the way data is maintained. Also, once the medical staff becomes accustomed to rapid image accessibility, it will be difficult to manage staff frustration from a long wait to retrieve images should there be a disruption in service. Many facilities are weighing alternatives, including complete spinning disk redundancy, or near line backups that are readily accessible to address such concerns. Providing redundancy for a centralized enterprise archive can be a far simpler and less expensive solution from managing the disaster recovery requirements of individual service area systems.
The Bottom Line
Any facility that has multiple service areas with digital imaging requirements needs to be contemplating enterprise image archive approaches. I’d very much like this blog to be an interactive forum for discussion of differing viewpoints and perspectives. Some potential questions I’d like to hear from you on include:
§ How important is this issue in the scheme of IT priorities? Which service areas have the highest priority?
§ What type of justification is necessary to approve?
§ How many facilities have multiple service areas doing disparate PACS solutions?
§ Who agrees or takes issue with the context of the enterprise image archive viewpoints expressed above?
§ As an IT executive, how comfortable are you in understanding imaging needs?
I look forward to your comments!