I have been thinking a lot about tools recently, in part because my "honey-do" list seems to be growing, but mostly because of my conviction that our informatics is, in fact a tool set, whose right to exist is limited to its ability to help us better render pharmaceutical care to our patients.
In the general case, I see tools as falling into two basic categories: hand tools and power tools.
Hand tools are small, and are generally useful for small tasks that require delicacy, or need to be performed in tight quarters in which power tools likely do not fit. I generally like to use my power drill as a screwdriver, but have learned that there are places where only a small stubby screwdriver will fit, and that there are some jobs (like assembling a desk from a kit) where the use of the power drill as a screwdriver is too hard to control and often results in my damaging the desk.
Hand tools tend to lack precision (primarily because they rely on the precision of the human hand - which is not all that great), and are limited in their scope by the physical strength of the user. If I want to dig a small hole or even a short and shallow trench, a shovel is likely just fine. If it needs to be long and straight, I will get the job done faster and more accurately with a ditch-witch (think of a chain saw that cuts dirt), or a back-hoe. If, as part of the process, I have to break up and remove 100-lb rocks, or if the ditch has to be over about 5 feet long and straight, I pretty much need to use a back-hoe.
If I am cutting a board, I can use a hand saw. The cut will likely not be very precise, and I will likely get tired very quickly if I need to make too many cuts. If I need to make a lot of cuts, I can do that better with a power circular saw. If I need to make those cuts very accurately, I would likely need to use a table saw or a power miter box. If I need to cut a bunch of boards and have them all come out exactly the same, I need a table saw or a power miter box. If I need to make a picture frame that has depth (the cuts therefore need to be angled in at least two dimensions) I must use a power miter box.
When I use power tools, I must use them differently than hand tools. If I try to cut a board with a circular power saw by moving it back and forth across the wood the way I would use a hand saw, I am unlikely to get a cut at all, though I might scar the wood pretty badly. I need to understand the best (and safest) way to accomplish my task with the power tool and use it as it was intended to be used.
That means that I have to be properly trained to use a power tool accurately and safely. A pneumatic nail gun in the hands of an experienced roofer gets roofing done quickly with each nail fully and properly embedded in the sub-roofing. That same tool in the hands of a 5-year-old is a recipe for a trip to the emergency room.
By analogy, our use of technology is, essentially, the use of power tools.
Our paper forms are hand-tools. They are pretty good at collecting hand-written information that we don't ever really expect to have to query or look at again, and they represent a good way to do record keeping on new processes until we understand those processes well. They are infinitely adaptable (you can write pretty much anything you want anywhere on the form in pretty much any language you want), but they cannot, and do not permit any constraint or standardization of the information they contain (other than by human compliance), they can seriously challenge any data recovery, and are easily lost or misplaced.
Similarly, in our roles in which we attempt to ensure that the correct medications are correctly selected, prepared, and dispensed, our "hand-tools" are principally our eyes. You don't have to read very far to realize that human cognition is actually evolved to not do well at this job; our tendency to get tired and our biases get in the way. This is not something you can train the brain out of; it is the way the brain is designed to function. Seth Hartman did an excellent presentation on this a couple of years ago at the Informatics Institute at the ASHP Summer Meeting. Google "human bias" and you will get a wealth of information on biases that govern our cognitive process.
Studies on human reliability indicate that, depending on circumstance, human error rates range between 1% and 25%. The higher the cognitive load (or stress) is involved with the work, the higher the human error.
Indeed, because our hand tools are so "flexible", it is not uncommon to find that the way processes are performed within our pharmacy enterprises vary widely by individual or shift, making those processes difficult to manage or improve.
Our power tools use technology to help us overcome our limitations:
- They permit us store large amounts of information in ways that make it easily retrievable
- They can be designed to constrain the way data are entered so that it is consistent and accurate
- They can sift through a large amount of data and find the "needle" that is data we need to see in the "haystack" of mostly normal data
- They can warn us when we break rules we don't want to break.
- They can positively identify drug products more reliably than our observation can
- They can lead us through complex processes in organized ways so that work is done consistently according to best practice
- They can capture detailed information about processes as a by-product of performing those processes that we likely could not capture by hand and still get our work done (if we remembered to capture it).
Just like other power tools, using these power tools requires that we understand how they are intended to be used. This means that we do not layer them on top of manual procedures and expect them to just make the manual procedures better. We have to know how best to use the tools, and adapt our processes to make the best use of the tools.
Just like other power tools, incorrect use of our power tools can result in unanticipated consequences. Bar code scanning, for example, can provide significant benefits in ensuring that correct medication products are used, but only if the database of scan codes is kept current, and only if the bar codes being scanned are actually on the products they are supposed to represent.
What this means is that when we apply these power tools to our practice, it is highly likely that our practice needs to change. While this is likely a bit painful (we tend to equate habit with competence), properly adapting the work to the tool set will maximize the value received from that tool set. Indeed, there is a lot of experience to indicate that layering automation atop a manual process (especially as variable and as poorly-understood as those processes tend to be) is a recipe for failure.
This, in turn, means that automation is a journey, not a destination. It means that automation needs an infrastructure of thought, and design, and best practice within which the tool set is used. It also usually means that there is an ongoing maintenance burden both of the technology and the process to ensure that the entire system is operating properly. It also means change for the workplace, with all the change management that implies.
So, let us not be incredibly surprised when technology that is just dropped into place fails to deliver on its value.Let us rather implement our technology thoughtfully, with careful planning both for implementation and ongoing maintenance, and let fully support our staffs during the change that such adoption must necessarily require.
What do you think?
Dennis A. Tribble, Pharm. D., FASHP, Ormond Beach, FL, DATdoc@aol.com
The ideas expressed in this blog are my own, and not necessarily those of my employer or of ASHP