Emergency department tracking board design that works
Every emergency department has a tracking board. Rather fewer have one people read during a bad shift, which is the only shift where it matters.
An emergency department tracking board earns its place when a charge nurse can see, in one glance, who is waiting, who is waiting on something, and who is stuck. That means fewer columns than most builds carry, status fields staff will actually update, and the boarding count in plain sight.
Emergency medicine is the only setting where the record has to work at a glance.
Everywhere else in a practice, a clinician opens one chart and reads it. In an emergency department a charge nurse has to hold thirty patients in view at once and decide, in a few seconds, which one has stopped moving. That is not a charting problem. It is an information design problem that happens to live inside an EHR.
The volume behind it is not small. NCHS recorded 155,398 thousand emergency department visits in 2022 — about 155.4 million — at a rate of 47.3 visits per 100 persons, with a median wait of 16.0 minutes to see a physician, nurse practitioner or physician assistant against a mean of 38.1 minutes. The gap between that median and that mean is the whole subject of this post.
Patient flow is what the board is actually showing
A tracking board looks like a list of patients. What it is really displaying is patient flow — the movement of people through a sequence of steps, and the places where that movement stops.
That distinction changes the design. A list answers “who is here.” A flow display answers “where is the queue.” The second is the question that gets asked at two in the morning, and a board built as a list will not answer it however many columns it has.
The practical version: every patient on the board should be visibly in one of a small number of states, and the time they have been in that state should be legible without arithmetic. Everything else is supporting detail.
Boarding: the number the board has to make visible
Boarding — holding admitted patients in the department while they wait for an inpatient bed — is the single biggest consumer of emergency capacity, and it is the number most boards bury.
AHRQ’s guide for hospitals on patient flow puts the scale of it plainly, noting that nine out of ten hospitals report holding or boarding admitted patients while they await inpatient beds. The same guide records that in 2007, the most recent year of data it cites, 1.9 million people — 2 percent of all emergency visits — left before being seen. Those figures are dated, and we are citing them as historical context rather than as today’s position.
What is not dated is the design lesson. If boarded patients are shown in the same colour and the same shape as patients still being worked up, the board is hiding the constraint it exists to reveal. Give them their own visible state and a count.
- A board that cannot answer “how many beds are occupied by people who no longer need us” is missing the number that determines everything else on the screen.
Throughput measures and where the clock starts
Two federal measures shape how emergency throughput gets reported, and knowing what they count changes what you put on the board. CMS defines two emergency department throughput measures: ED-1, the median time from emergency arrival to departure for patients admitted from the department, and ED-2, the median time from the admit decision to departure. Each is stratified three ways — all patients, observation patients, and patients with a psychiatric or mental health diagnosis.
ED-2 is the boarding measure in disguise. The interval it captures starts the moment somebody decides to admit and ends when the patient physically leaves, which is precisely the period during which the emergency department is holding a patient it cannot treat further.
The operational implication is that the admit decision time has to be captured as an event, not inferred later from an order timestamp. If nobody records it deliberately, your ED-2 figure is a reconstruction and the board cannot show the clock at all.
Status fields that people will actually update
Here is the rule that decides whether a board stays accurate: a field that requires someone to remember to update it will be wrong by the third hour of a busy shift.
So derive what you can. “Labs pending” comes from the order and the result, not from a checkbox. “Imaging in progress” comes from the modality. “Disposition decided” comes from the admit or discharge order. Each derived field is one less thing competing with patient care for a nurse’s attention, and it is right without anyone maintaining it.
Reserve manual status for the small set of things the system genuinely cannot know — usually “waiting on a consultant” and “ready to move.” Two manual fields that are always current beat eight that are sometimes. That trade is the core of rebuilding the workflow around the board rather than adding to it.
Not sure your board is showing the right five things?
We watch a shift, map what your board shows against what your charge nurses actually ask, and hand you written findings.
Door-to-doc, and the trap of a single average
Door-to-doc — arrival to first clinician contact — is the metric most departments watch, and a single average of it is close to useless.
Look again at the national figures: a median wait of 16.0 minutes against a mean of 38.1. A distribution where the mean is more than double the median is not a distribution with a typical case. It is a distribution with a long tail, and the tail is where the risk sits. Reporting the mean tells you almost nothing about the patients who waited two hours.
Show the median and a count above a threshold your department sets, and show it by hour of day. In our experience the pattern that emerges is not random — it clusters around shift change, a particular arrival window, and the hours when boarding peaks, which means it is addressable rather than inevitable.
What to fix first
- Cut the board to what changes a decision. If nobody acts on a column, it is costing attention and returning nothing.
- Capture admit decision time as an event. Everything about boarding depends on it.
- Derive every status you can. Manual fields decay fastest exactly when accuracy matters most.
- Report the median plus a tail count. Never a lone average.
None of this needs new software, which is the point. It is configuration, event capture and a decision about what deserves screen space. Qventive provides emergency department tracking board services across New Jersey as part of our emergency medicine EHR and IT work and, where the questions run wider than the board, a broader emergency medicine practice management engagement.
One caution worth stating plainly. A tracking board is usually the most publicly visible screen in the building, frequently mounted where patients and visitors can see it. That makes its contents a HIPAA risk analysis question as much as a design one, and it is a question to settle before the screen goes on the wall.
Tracking boards: common questions
How many columns should a tracking board have?
Should the board be on a wall display or on workstations?
What are the HIPAA implications of a visible tracking board?
Can we automate status without changing how staff work?
How do we know whether a board change actually helped?
Let’s meet
Qventive works with emergency medicine groups across New Jersey on board design, event capture and throughput reporting. Thirty years healthcare-exclusive, from our office in Hackensack.
- 155,398 thousand ED visits in 2022; 47.3 visits per 100 persons; median wait 16.0 minutes, mean 38.1 minutes. cdc.gov
- ED-1 and ED-2 throughput measures and their three stratifications. cms.gov
- Nine out of ten hospitals report boarding admitted patients; in 2007, 1.9 million people (2 percent of visits) left before being seen. ahrq.gov