Healthcare providers produce an ever-expanding collection of electronic health records and medical imaging data.
Storing such a high volume of data can be challenging, and this is compounded by long-term retention requirements. As such, healthcare providers often search for creative solutions to data storage problems.
Here are some of the storage options I've seen in use by healthcare organizations:
- Storage area network (SAN)
- File server storage
- Network attached storage (NAS)
- Tape storage
- Cloud storage
- Robotic tape library
Most large healthcare organizations make use of a storage area network (SAN). Portions of the SAN are often used to house virtual machines, while other areas of the SAN are used for data storage. This means that plenty of demands are already being made against the SAN well before the long-term retention of exponentially growing data is even brought into the equation.
The nice thing about SAN environments is that they deliver high performance and they are very flexible. It is relatively easy to expand a SAN as an organization's storage needs increase. The problem with doing so, of course, is that SAN is expensive.
In any large healthcare organization, there is always some data that undoubtedly needs to be stored using a SAN. That data might need to be protected by the various forms of redundancy provided by the SAN, or there might be a benefit to storing the data on high-performance media. For such data, there is a legitimate business need for using SAN.
Oftentimes, however, there is data that really doesn't need to be stored on a SAN. Most hospitals, for example, have static data that is rarely accessed. It doesn't make sense from a business perspective to use an expensive SAN for data that could just as easily reside on a less expensive form of storage. Healthcare organizations typically address the challenges of long-term data storage by classifying data based on its age and the frequency of access. The data is then moved among various storage tiers according to threshold values set by the administrator.
Alternatives to SAN for storing seldom-used data
There are a number of different ways that static or seldom-accessed data can be stored. Some organizations choose to use file server storage or network attached storage (NAS), while others depend on tape or cloud storage. It is becoming increasingly common, however, to use some combination of these storage types.
To give you a concrete example of how storage tiering might be used, consider the hospitals that I used to oversee. At the time that I worked there, the organization had a policy of keeping each patient's medical records for 10 years after the patient's death. Through usage tracking, it was determined that if a patient died in the hospital, their medical records were accessed on a fairly regular basis for the next couple of months. After that, the records were almost never accessed again.
This is a perfect example of static data that must be kept, but that may never be accessed again. You probably don't want this type of data taking up space on your SAN, so where should you store it?
One solution is to archive the data to tape. While this is certainly a viable option, it must be done carefully. Sometimes, even the most mundane data can unexpectedly become important. Suppose, for example, that eight years after a patient's death someone completely unrelated to the patient decided to file a malpractice suit against the doctor? Attorneys might subpoena the medical records of everyone that the doctor had treated for a similar condition. If this sort of thing should happen, you really don't want to have to wonder where the tape containing that archived data might be. The data needs to be readily accessible.
There are a few different ways of dealing with this problem. One option is to archive data to the cloud instead of to tape. Although cloud storage is certainly a viable option, it can be expensive. Most of the cloud storage providers bill customers monthly for the amount of storage space used. In other words, as the volume of data that you are archiving grows, so do the monthly charges from the cloud provider.
Another option might be to use a robotic tape library. One of the organizations that I used to work for had a large data warehouse filled with tapes. If archived data was needed, a robot would automatically retrieve the tape from a shelf and insert it into a tape drive in the next room. This type of storage worked well because the cost per gigabyte was low and there were no ongoing storage fees. Of course, the initial investment in hardware can be quite expensive.
Whatever type of storage you decide to use, you will need data life-cycle management software that can automatically move data from one storage tier to another as the data ages. The software should also be able to automatically purge data when it is no longer needed.