Counting the Ways to Solve the Counting Problem w/ Hospital Supplies
Since moving to San Antonio in May, I'm struck by what a green, pretty city this is. Pictured at left was a day last week when I had a chance to walk from the co-working space and community Geekdom to the convention center… via the Riverwalk. It was about 95 degrees, but it's pretty shady down there.
I was invited to be a guest of a company, Spacesaver, to visit the exhibit floor at the AHRMM 2012 show – focused on resource and materials management for healthcare. I got to spend an hour looking at their solutions as well as two of their competitors… how to improve healthcare supply replenishment systems?
Disclosure: Spacesaver provided a free guest pass for me attend for an hour, but I have no other current financial connection with the company.
A common theme at three of the vendors I saw touched on something I've talked about a lot — the hospital “par level” systems for restocking materials involve a TON of waste — the time spent counting each item.
For those unfamiliar with par systems (say, those of you in manufacturing, where I'm pretty sure this approach isn't used… I had never heard of it), let's say the par level is 100 units. Every day, a materials management person might count (or estimate) the items remaining. If there are 60 remaining, they bring 40. If there are 80 remaining, they bring 20.
The counting can be very time consuming. There's also time required to carefully bring that quantity that's required for a certain day. If you need to bring 20 and the items are packaged in boxes of 25, you'll either just bring a box of 25 or take 5 units out to bring just 20… waste! It's a “pull system,” but it's a very labor intensive pull system.
Here's a great video from a Pennsylvania hospital about converting from a par system to kanban, a different pull system that's less labor intensive.
A two-bin kanban system makes things very visual, simple, and binary. If the par level was 100, we'd set up a two-bin system, with two bins containing 50 units each (assuming the par level had been set correctly, but that's a story for a different day). Again, any basic, manual two-bin system eliminates counting… whether you have any technology involved or not.
There's agreement that “counting is bad” (as shown in the displays I photographed) — but there are different countermeasures being proposed by the different technology and systems vendors, as shown below.
Spacesaver has a two-bin kanban system solution where the technological innovation is an RFID tag on each shelf. One one bin (or one half of a single bin) is empty, a staff member takes the card from the front of the shelf and puts in what looks like a printer with a slot (which is a device from a company called LogiTag). That slot/box reads the RFID tag and tells the materials management department what needs delivered. Spacesaver calls the system StockBox.
The RFID tag is meant to be less time consuming than scanning bar codes that might be printed on bins or on kanban cards. There's, of course, the extra cost of the RFID tags, the StockBox readers that you would need in each stock location, etc.
A company called PAR Excellence is also against counting… and scanning, apparently.
Their technology solution is to have a scale attached to where each bin hangs on the wall, so the weight of a bin is constantly measured (which is translated into a number of items remaining, based on said weight). Read more about their solution here.
This approach promises to require less effort than dropping an RFID card into a hopper… but it also requires the cost of scales, the scales have to be on the network, and there's a lot of data you would have to maintain.
Another company, Logi-D, is opposed to counting (see below):
They have an RFID approach that's similar to Spacesaver, but instead of putting cards in the hopper / box, you put the tag onto an electronic board that hangs on the wall. That board gives visibility into what's waiting to be replenished in a way that's different than Spacesaver (which gives you visibility on the bins themselves…. no RFID tag, you see a red label, showing that it needs replenishing).
So there's two different approaches to RFID and a scale-based system.
Logi-D has another new system that has a “gee whiz, that's cool” factor to it… but Lean thinkers don't choose any technology because it's new and cool. We choose what's effective (based on the Toyota Way principle that says we should use only “reliable, throughly tested” technology that “serves our people and our processes.” In other words, don't be the first to use a new technology… but don't be afraid of technology if it helps.
Anyway, the new solution from Logi-D, called VC-iD, is based on a computer vision system. Yup, a camera looks down at every stock shelf. You can read about the RFID and VC-iD solutions here (and see a video of the VCiD system). When a bin in a drawer is empty, the camera and computers “see” the bottom of the bin and register it as empty. As with the other company's system, this information is sent electronically and is used for inventory records and/or restocking.
The VC-iD seemed to work, in the live demos…. but it seems like the most expensive way of address the problem.
What was the original problem? The problem was that a par level system requires a lot of labor to count and restock.
I asked a Logi-D rep “isn't that a really expensive way to solve a fairly simple problem?” and he responded that “well, it depends on the value of the items being restocked.” I guess you might use a system like that for expensive implantable items instead of gauze and bandages in a utility room.
What are your reactions to systems like this? Helpful technology that solves a real problem you have? Technology looking for a use?
How are you communicating your kanban replenishment signals? Computer? Fax? Physically carrying the bins? Bar code? RFID? Carrier pigeon? There are lots of ways to solve a problem…
What do you think? Please scroll down (or click) to post a comment. Or please share the post with your thoughts on LinkedIn.
Don't want to miss a post or podcast? Subscribe to get notified about posts via email daily or weekly.
These technologies have always seemed like overkill to me. Just like with performance measures, the most useful and utilized systems seem to be simple and manual. However, they also are the hardest to start because they require real management and leadership, rather than an expensive piece of technology that people can rally around. It would be great to hear from anyone that has one of these types of systems what their actual experience has been – are stockouts lower? Is total inventory onhand lower? Is it easier to maintain overall than a manual system?
I agree that I’d prefer simple, manual systems.
You’re asking great questions about what the benefits are, if they are there. I didn’t have that in depth of a discussion with any of the vendors.
Before putting a technology like this in (even simple bar codes), we should ask “what is the problem we are solving?” or “what is the problem with our manual system?”
I’m unsure why hospitals are so in love with their par systems. It is essentially taking a physical inventory of the whole hospital every day (if its done correctly). No other industry does that. The electronic options presented here are much more complex than necessary. A two bin (or other kanban) system can signal resupply needs more efficiently and be made to work with whatever electronic material handling systems they already have in place.
I’ve seen very effective use of simple kanban at several hospitals. The result is fewer stock outs, fewer expired supplies, happier material handlers, and happier front line staff.
Yeah, I had never heard of a par level system before getting into healthcare. Not used in manufacturing – at least what I saw in different industries.
I’ve seen kanban systems be very effective with manual communication (carry cards or bins to the stockroom) or simple bar-code ordering through the hospital’s materials/financial system.
Although I recognize that LEAN practitioners are looking to solve manual problems with manual processes, automation ensures less process waste or “MUDA”. A simple process cost vs technology cost calculation and cost vs value will answer if it makes sense.
As an example, the time consuming and error-prone par level counting is both labor intensive and the value output is rather low due to the inaccuracies it generates which drives stockouts and sub consequent “just-in-case” inventory levels as a counter-effect.
RFID-enabled Kanban releases resources in the demand chain that can be redeployed to more value-adding activities across service lines and the minimum human intervention provides higher accuracies, preventing consequent MUDA across the front line of service.
Kanban itself is a superior practice to par level, but with RFID, the demand is captured in real-time and resources are released from the demand chain which makes the technology cost highly justifiable with ROIs typically within a year.
So the question then becomes “what is the monthly labor cost vs technology subscription of managing the SKU and which has the highest cost2value ratio?”
I think the goal is solving problems with the simplest, most effective approach possible. Does “real time” mean this very minute or within the day? I think, except for very expensive supplies, that there are diminishing returns from being “more realtime.”
I agree that looking at the cost vs benefit is the key. I’m afraid some of these technologies would be more expensive than necessary.