Silos: Great for Grain. Not for Clinical Data.
I talk to lots of hospitals. In fact, it’s pretty much all I do, as they are my customers and I’ve been doing this for about 20 years now. Conversations often go like this:
Me: “So what storage platform do you currently use to store your critical data?”
CIO: “Well that depends – you see we store our MEDITECH data on an ABC Company array because that’s what our MEDITECH systems integrator recommended for performance. We store our PACS data on an XYZ array because that’s what our PACS vendor recommended for reliability. We also have a VDI implementation that we store on an XXX company array, because it has flash. And oh, we keep all our user shares on a ZZZ array because it was cheap.”
Me: “How do you manage all these different systems?”
CIO: “We have different teams for each.”
Every hospital I speak with has AT LEAST 2, if not 3 (and sometimes more) different arrays, usually from multiple vendors, and a team specializing in each. Common sense says:
- “Why?” Wouldn’t it be simpler to manage a single array?
- Wouldn’t it lower costs and energy and space and be easier to maintain and service?
- Wouldn’t it require less staff time and management energy, and allow everyone to focus more time on the clinical users?
The answer to all those questions has to be “yes.” But we’ve fallen into an unintentional silo trap that’s been set by the software providers and many of the storage array vendors. As my esteemed HP colleague Jorge Maestre, recently blogged, “Every vendor focuses on what they do best, and as a result we’ve ended up with data centers that are a complete mish-mash of solutions.” It requires a greater level of resources to effectively manage so many products. How many IT departments have seen an increase in their resource pools lately? Not many. Indeed, the constant refrain continues to be “we have to do more with less.”
What’s to be done? Let’s start by looking for ways to eliminate those silos. I have yet to speak to a hospital that wasn’t at least intrigued, if not totally enthralled, with the prospect of running ALL mission critical apps on a single, scalable storage platform. Siloes are useful on the farm, but in the data center the technology exists to free yourself of silos and consolidate your data. It’s cost effective and readily available.
Let’s consider HP’s 3PAR family of storage arrays. Most major EMR and PACS vendors support and highly recommend 3PAR. 3PAR was designed for multi-tenant environments, managing the data in such a way that space is easily re-allocated to those apps that need it. A major benefit is you don’t have “idle” space for one app while another app goes starving until that new shelf of disk drives arrives.
Recently I heard an interesting story. The hospital’s EMR vendor indicated that a dedicated array was required. But, it turns out that the array proposed couldn’t handle the additional load presented by the other applications. The same EMR vendor had no such restrictions with the 3PAR we proposed. The moral of the story: you don’t have to sacrifice performance or support to gain the efficiency and cost reductions hospitals need to achieve.
Logically, the next thought is “Wouldn’t more arrays reduce the probability of having a key application go down?” Great question, but not so fast…HP 3PAR can provide up to SIX 9’s of availability. Sounds like a lot. I did the calculation and it equates to 31.5 seconds of downtime PER YEAR. That’s both unplanned AND planned! So you don’t have to sacrifice reliability a bit, in fact you’ll improve it to the level required in today’s digital hospital.
So my advice is to do your homework, fully explore your options, stick to your guns and make your data center easier to manage, more efficient and less costly. Leave the silos to the farm and consolidate your storage – you’ll be setting the foundation for a strong, efficient, and “enterprise” storage strategy.
Contact us for more information about a Comport Healthcare solution for your environment.