When it comes to imaging infrastructures, healthcare providers are facing numerous challenges and it is tempting to transfer responsibility of IT activities with their associated risks and burden to a third party. Subscribing to a managed service to store, share and distribute healthcare images from a hosted platform sounds like a good alternative. But will it be more expensive than if the organisation does this itself? How does this relate to increasing cost pressures in the healthcare sector?
by Bernard Algayres
Globally, the volume of digital information is exploding. Recent reports from the IT market intelligence provider, IDC (March 2008) show that the volume of digital information produced will multiply by 10 between 2006 and 2011. Another IT advisory company, Gartner is expecting a 69% annual capacity growth in this period, an even higher growth than was expected earlier.
Healthcare data follow the same trend, with the volume of information either produced digitally
or digitalised increasing. This is particularly true in imaging departments with several cumulative phenomena contributing to growth. New generation CT and MR machines are producing a larger number of heavier images. The volume of images per exam is therefore so large that new tools are necessary to improve diagnostic efficiency. At the same time these CT and MR technologies are more popular, and increasingly used for diagnosis.
Retention time also impacts on volume: legal requirements have evolved making it now critical to be able to present images in case of requests from the authorities.
Collaboration is now the new way of working in all industries and this is also true in healthcare: patients often receive critical care from multiple healthcare facilities (e.g. chronic disease treatment requiring coordination of different specialties). As a result, patient data may be scattered among various clinical systems across multiple sites with no simple access to aggregated historical records. The lack of healthcare professionals in a number of specialties also increases the need to easily access patient information from any location.
Demanding healthcare professionals
Healthcare professionals are under pressure: they need to provide the best care in a reduced- cost environment. Access to the latest clinical tools and optimising workflow is therefore key. In addition to their traditional activity of setting up IT infrastructures, hospital IT professionals need to act as real business consultants, understanding the needs of healthcare professionals in order to find the right solutions.
The cost impact
Long-term preservation of information is one of the most challenging parts of the hospital IT organisation. Healthcare organisations trying to find the best solution for keeping data available over a 10 to 15 year period of time often consider investment as the main portion of their costs. A number of studies contradict this. Gartner estimates that capital costs count for less than 20% of the Total Cost of Ownership (TCO) of a storage solution. When dealing with an archiving infrastructure, we can even consider that the proportion is lower, and that the costs of maintaining information accessible during tens of years are largely underestimated by IT organisations.
Let us consider a classical archiving infrastructure and list the mandatory tasks required to maintain the minimum service levels required by users.
Hardware and software maintenance
This is provided by the technology suppliers and usually covers problem fixing in a reactive mode with on-site support and software updates.
Daily infrastructure monitoring
In order to make sure that the system is working properly, daily monitoring is important.
- At the hardware infrastructure level: this means insuring that no hardware component such as a RAID disk or a tape library is damaged.
- At the application level: all the processes that run the archiving application must be working properly, particularly management of the database that is the kernel of the system. This task requires a very good knowledge of the archiving application itself.
- At the data level: it is critical to ensure data protection through the backup or replication processes, data security through the analysis of access logs to the data, and data integrity. This part is key to guarantee long term preservation.
This covers all the tasks related to the occurrence of a problem, including:
1. Problem detection and identification: identification of the exact source of the problem
2. Escalation process management: having the right supplier handling the problem with the proper level of priority.
3. Resolution management: implementation of the fixes by the supplier, testing and acceptance of the fix, and communication with the users impacted by the problem.
4. Hardware and software update/upgrade management: although software upgrades may be part of a supplier maintenance contract, the hospital IT has to manage the
- Management of the supplier intervention
- Management of the user information for possible downtime
- Minimum testing and acceptance of
Given the exponential growth, hardware and software provisioning must be addressed regularly. It may require a re-design of the infrastructure if not properly designed at the beginning, or when the infrastructure reaches its capacity. Provisioning covers:
- New capacities: costs here include both new investments and management costs to minimise impact on users, and to communicate with them. If not anticipated, it can also lead to downtime costs.
- New applications that need access to data: during the life of an archive, new applications such as patient record systems, new administrative systems or new clinical tools will need access to data.
This is certainly the most difficult part and will become increasingly complex over time. As we have all experienced, technology renewal rate in IT is usually between three to five years. Knowing that retention time of data is close to 20 years, it means that each component of the infrastructure will require changing four to seven times during the lifetime of information, with possible associated migrations. As soon as obsolescence occurs different costs have to be taken into account:
- Re-investment: the system has to be renewed and new systems must be integrated in
- Migration costs: all ingested data in the archive need to be moved to the new system. Over time, and as volumes of information grow, migrating data can become a real issue.
Figure 1 illustrates the combination of obsolescence and volume of data, and the complexity that will increase when volume grows.
The Managed Services path
As we have seen, the healthcare IT environment is becoming more and more challenging, with data volume growth and increasing information access requirements. At the same time, the management of the clinical data repository necessitates more and more resources, with high level expertise in storage technologies, database management, network and security.
How can managed services be cost effective?
The proactive monitoring approach enables the early detection of potential problems before any impact occurs on production. Figure 2 gives an illustration of this.
In the standard break and fix service model, problems are detected by users (white curve), usually when it is too late and system operations are seriously degraded. During this period, it takes time to analyse the problem and investigate root causes, and system down time usually happens where situations may not be fully recoverable, leading in extreme cases to data loss. This is typically the case for systems resource overloading or exhaustion: disk reaching limits, no more tape for backups or server overloaded by migration.
Thanks to Managed Services monitoring resources and tools, problems can be detected in a very early phase (green curve) by the monitoring centre and before there is any impact on users. Problem detection in an early phase makes it possible to solve such problems faster and more efficiently. In most cases, proactive monitoring can even prevent problems from occurring.
Technology is the basis of a reliable and cost effective service offer. It is designed for service and fulfills the following criteria:
- Flexibility: the technology must be able to handle different configuration mixing both on site and off site storage.
- High availability: this is what will allow the supplier to guarantee service levels.
- Multidimensional scalability: the technology must be incrementally scalable in capacity, performance, storage technologies and application connectivity. This incremental
scalability is key for cost efficiency.
- Enhanced connectivity: the solution must interconnect through different protocols to different applications from different vendors.
- Clinical information lifecycle management: this is particularly important in order to optimise cost and ease migration.
Providing a managed services offer requires a dedicated organisation, which must employ highly skilled experts who are always available to make sure that the system is optimised in terms of availability, security and performance. Advanced monitoring tools are also key in order to ensure the quick detection of any issue, the immediate preliminary diagnostic, and quick involvement of the right level of expertise.
We have seen that analysts consider that in the Total Cost of Ownership of a classical storage infrastructure, the investment costs are less than 20% of the total cost. This means that a $200k infrastructure will finally cost $1M in a five year period plus migration costs. Experience shows that these costs are under evaluated.
Based on different studies across different environments, estimation is that the cost for migration is around $5 000 per terabyte and that due to obsolescence, 25% of the volume is moved each year. This shows that the total obsolescence costs can become a real problem. A hospital, say producing 250 000 annual studies for an equivalent of 25 Terabyte of uncompressed data/12 Terabyte compressed, would incur yearly migration costs of around $100 000 at year seven, $150 000 at year 10 and $225 000 at year 15. This excludes the volume growth for this specific example.
Archive systems are complex systems that combine servers, databases and a variety of storage subsystems, often over several locations. They must be maintained over long period of time. Hospitals often consider managed services as a way to offload their IT department and transfer storage management tasks to a third party. However, when looking at the cost aspects, they are usually surprised that their real internal costs are much higher than thought, and that managed services are a way to save costs.
In addition, managed services reduce the risk of system unavailability or data loss thanks to permanent monitoring. Also, collecting information from different sites to be stored and preserved on a datacentre broadens access and distribution of images as these activities can be achieved direct from the datacentre with enhanced security, instead of opening each hospital site to the external world.
Rather than investing in infrastructure, healthcare is increasingly following industry trends by subscribing to services on a pay as you go basis with service level agreements. Carestream Health with its eHealth Managed Services, is at the leading edge of this movement.
eHealth Managed Services