- History, How Bad Design Kills, Posture and Metaphors
- Interruptions, Memory and Situational Awareness
- The Magic Number 7 (or maybe 4?), Forced Errors, Triage, and Color-Blindness
- Where is my [fill in the blank]? and Passive Tracking
- Information Design 2
- Lessons from Tufte
- Natural Mapping, Search and Affordance
In its most pedantic definition, an ED tracking system is a system for tracking things in the ED. Tracking things that, well, things that we need to keep track of. People. Orders. Labs. Consults. Messages from docs. Many different bits of information that are needed to keep the ED running smoothly. One of the reasons for such a system for keeping track of things is that the human mind is limited in the number of things it can keep in short-term/working memory. And unlike humans, computer are great at remembering things. (Not so good at figuring out what do do with this information, or, as is evident from many of the posts on this blog, currently not so good at presenting information to humans.)
Anyone who has attended a standard adult-learning course (e.g., first aid, CPR, ACLS or PALS or ATLS instructor course) will remember one of those principles of adult learning: we can only keep “X” things (I have been told four, five, three, and seven) in short-term memory at once. We’re told to limit the text bullets on our overheads, slides or PowerPoint screens to five.
This actually has some basis in science.
The limits on short-term memory (or working memory, to use a more modern term; the links provide the pedantic differences between the two) were popularized in a landmark study by Miller in 1956, entitled The magical number seven, plus or minus two: Some limits on our capacity for processing information. But even as far back as 1905, Wilhelm Wundt, a physician trained at the University of Heidelberg, and known as the father of experimental psychology, had established a similar limit (Wundt’s Limit). More modern studies suggest a smaller number, about four, for the number of similar items we can hold in mind at one time. A recent study suggests that short-term memory decays quickly over time, especially if, as Wundt observed, competed with by other sensory processes.
In the ED context, for “competition by other sensory processes” one can cite screaming children, perseverating demented elderly, psychotic patients in restraints, and those intoxicated with a variety of substances, not to mention urgent demands from other patients and hospital staff. A recent paper, Emergency department workplace interruptions: are emergency physicians “interrupt-driven” and “multitasking”? points out that task interruptions have been shown, in other contexts, to contribute to error – and we may reasonably attribute at least part of this due to evaporation of information from working memory.
From the above paper:
A forced error is one in which task demands exceed physical capabilities or resources available. For example, an emergency physician is sometimes forced to make a decision on the basis of inadequate patient information. Instant information about the patient’s medical history, allergies, and medications would decrease the incidence of these forced errors.
Thus, the perceived need for some system, a system outside of the working memory of those who are being interrupted, to keep track of the critical bits of information: an ED tracking system. It is also true that not knowing what is actually happening with an emergency physician’s patients in the ED may be a source of “forced error” – if one doesn’t know that the x-rays are done, one doesn’t know to look at the x-ray, one then doesn’t know the patient has a pneumonia, and one doesn’t know to order the antibiotics, delaying time to antibiotics. And, once the emergency physician orders antibiotics, if the nurse doesn’t know about the order, that further delays the time to antibiotic administration, which at least from some studies seems to have an influence on mortality from pneumonia. And, the Harvard Medical Practice Study says that, of errors in the ED, over 90% were preventable.
There is a tide in the affairs of men,
Which, taken at the flood, leads on to fortune;
Omitted, all the voyage of their life
Is bound in shallows and in miseries
–Shakespeare, Julius Caesar, IV, iii, 217
Any emergency physician, nurse, secretary or tech can probably provide recent examples of how bits of information being temporarily lost has delayed patient care, or in some cases, adversely affected the health of the patient. I can vividly remember, from an emergency response one night while on-call as a resident, where failure of a cardiologist to see an elevated prothrombin time before a cardiac catheterization resulted in an avoidable patient death.
So, busier EDs need some help to keep track of things. At need, this can and has included information scribbled onto gauze-pad wrappers, the corner of a patient’s sheet, the leg of a pair of scrubs, or even someone’s hand. And indeed, experienced ED clinicians can and do use such “information first-aid” techniques when and if needed.
Such information first-aid techniques point up how smart humans can and do routinely compensate for broken systems in inventive and intelligent ways — something that seekers for “human error” seem to ignore, and something that those engaged in the scientific study of medical error relish as examples of how humans can compensate for “broken systems.” But still, many in the ED have thought there must be a better way. And indeed, over the years, as EDs have grown, prospered, and then become inundated with the failures of the rest of the medical system in the U.S., more formal information-tracking mechanisms have evolved. Although this article focuses on the use of computer technology for ED information tracking, it is worth spending a few moments discussing ED tracking in a more general sense, exclusive of computers.
And, to jump ahead a bit, one of the points of this series of posts is that today’s computer-based ED tracking systems are woefully immature in terms of their potential. This is not because today’s computer hardware is inadequate to meet the ED’s needs – it is because we have not yet developed software adequate to the task. And looking at how EDs have used non-computer technology to aid in ED tracking will aid us in selecting and developing computer software that meets the ED’s real needs.
The much-refined and extensively-validated 5-level Emergency Severity Index (ESI) is now almost universally used as an ED triage system in the USA. Once the nurse has established a triage priority, this – and the patient’s time order within the triage level – should determine who should be seen next by an emergency physician or mid-level practitioner.
Those with ESI 1 need to be seen right away to have a chance of survival. Those with ESI 2 should seen in the order in which they arrived, and after all those with ESI 2 have been seen by a practitioner, then those with ESI 3 should be seen, and so on.
I will say little about the practice of “reverse triage” – establishing a Fast Track in the ED to see the lower ESI patients before those with more serious problems – except to say that it only makes moral and ethical sense if you use cheaper, less-capable and less-expensive rooms and practitioners to care for these less-acute patients.
In order to make sure that emergency physicians and mid-level practitioners see patients in appropriate order, based on those triage priorities, EDs have derived a number of systems. Some EDs place a printed chart in a colored binder or on a colored clipboard with the more strident colors indicating a more-urgent patient. A common color code is shown above.
One problem with using colored binders, clipboards or tinting of a box on a computer’s tracking screen is color-blindness. About one in ten men have some element of color blindness – and anything this common can’t be considered a disease, but more a genetic variation. The figure shows what the ESI color scheme looks like to someone with the common deuteranopic red-green color blindness. (Courtesy of the free Photoshop Visicheck filter. This assumes you’re not color-blind. If you’re color-blind, I don’t need to be telling you about this.)
Another figure shows a tracking screen that uses the background behind the patient’s name to indicate the ESI number, using the color scheme noted at the beginning of the post.
As shown by the same figure in color-blind simulation, it will be difficult to use by 10% of men due to color-blindness.
A better alternative is to dedicate a column to the ESI, coloring the blocks in this column by ESI. While every pixel of a tracking screen is valuable, dedicating a column to something as important as ESI makes sense. A third figure, with an expanded tracking board, shows the use of a dedicated column for ESI.
Many larger EDs use pagers to summon a variety of clinical personnel for high-priority patients, e.g., major trauma, cardiac arrest.
Some EDs use the chart’s clipboard or binder color to indicate the patient’s location (e.g., red=Pod 1, blue=Pod 2, etc.) and simply place the charts in priority order, so the next physician who is free to see a patient will take the front/top, i.e., most urgent chart.
One nice feature for a tracking system, and one I’ve yet to see implemented, is to have a specific screen that tells practitioners which patient to see next. With a small ED as shown in the figure here, it isn’t hard to pick the next patient to be seen in proper triage order. But with a bigger ED, a dedicated function would be helpful. If one were to click on the ESI column header, a popup box (much like the way pictures in this blog pop up over the text when you click on a thumbnail graphic) could list the next few patients to see in proper triage/time order.
I will end this post with that tantalizing vision of the future (and clue to developers). Next time, will will speak of geographic tracking: where is my patient?
Tags: Computers, ED, ED Systems, Emergency Department, forced error, Healthcare, Healthcare IT, Human Error, Information Technology, Interruptions, IT, short-term memory, The magical number seven plus or minus two, Tracking System, triage, Tutorial, Wilhelm Wundt