Getting Real on Emergency Department Throughput

There are problems using throughput metrics as the primary source for decision-making both for Emergency staffing and throughput improvement engineering. 

Using metrics alone as the key source of decisions can quickly turn re-engineering into re-enginerring.

There is simply no substitute for direct observation study. If you are a decision-maker, you need to get reliable info first hand, too. That's you, not just a consultant.

Metrics / Schmetrics?

We have seen a recent trend over the last fifteen years in Emergency Services and that is posting waiting times on billboards, or the Hospital's internet front page that are very short - 20 minutes, 15 minutes, 10 minutes.

When we dig into the data, we find that the traditional standard measure of "Door to Doctor" and "Door to Treatment Begins" has been replaced with "Waiting to be Seen".

The numbers look great, but a great deal of actual patient waiting time has been left out of this measure.  

Median Times : a Mixed Metric

Another newer favorite change that obscures measurement of delay is using median length of stay and staging times rather than average times when tracking and reporting Emergency Services delay and length of stay. Median times ignore the longer waits. So they are much lower than the actual averages. And that feels good.

And median times are easier to improve. And that feels great.

But there is a dark side to median times...

You can eliminate bed holding altogether and this won't change your median patient time by much if at all.  

The median is just the longest length of stay of the bottom half of your patients. All patients here longer than the 50th percentage are ignored in the Median calculation.

If your median length of stay is 3 hours, and you move all your bed holding patients here 5, 6 and 7 hours down by a full hour to 4, 5 and 6 hours, your median won't change one single minute. Your dashboard will look the same, as if no change had taken place.

But it will change the average length of stay substantially. Hence the need to use average times if you are serious about your delay problems.

Then again, using medians provides a much lower figure than the average length of stay. And medians are highly sensitive to improving throughput of your middle length of stay patients.

Improving median times a lot only requires moving patients in the middle of the length of stay distribution who are just over the median to just under it.

And this has led to a focus on low acuity patients. That's a good strategy to start with. But when the total median length of stay drops folks plead "Mission Accomplished" and the real issues haven't been fixed.

Use Both Medians and Averages in Your Emergency Services Dashboard for Every Track and ESI level

Have your dashboard show both median and average length of stay and throughput times: Median times are more sensitive to the majority of your lighter acuity / mid-range length of stay patients in each acuity category; and the average delay and length of stay times are more sensitive to your longer-delayed patients in each of the same categories.  

The Average delay times will be the critical sensitive measures on improvements in boarder patients and longer stay outpatients. This will be the measure for effective teamwork with your inpatient bed management teams, units and care management outplacement  and transfer activity. 

Using median times alone not only hurts actual improvement work around serious delays, but also inadvertently short changes staffing.

Tracking these things isn't enough, though. Data must be interpreted.

But observations can get you there faster.

How can you take that next step in the journey of your own Emergency Department?

See the Flow - Be the Flow

If you want to understand you patients' experience, and the experience of the care givers and Providers a little better,  take some time every week, step away from the very large responsibilities of manager, director and VP, and step into the liberating role of  anthropologist / observer. Don't ask others to do this and report to you without having done this first yourself. Notice for yourself the stages of care individually, and the situation of your providers. And feel free to ask lots of questions...but be sure to spend the majority of your time observing.

If you don't have time in your busy work week, come in on a Saturday, Sunday  or a night shift, for a few weeks. Don't make it a big deal, like a gift you are giving your staff. Or a burden you must do because they did something wrong. Make it like a duty you have owed them for a while, and have been lapse in, and are just now catching up.

Just take an hour or two every week and watch. If you are the manager, director or VP, you won't wait for data to debate at team meetings in a far away conference room. You won't let the conversation devolve into stages and handoffs and questions about what really happened. That's so re-active! So last century!

 At some point you will intervene directly based on what you are seeing, when there is nothing to debate. You might start by asking some questions to help understand what you are seeing. The answers won't be "usually" or "I think normally". They will be what is happening now with this patient and their care givers and providers.

By observing first, you may learn something you didn't know, before you act. And act differently. 

The First Thing You May See...Folks Are Overworked During Busy Hours

You may see, right away,  that your staff is a bit overburdened, that using median or average length of stay values to derive Patient: RN ratios actually short changes your staffing. The actual ratios can peak well beyond the "average".

And you will also see that the presumed support systems upon which staffing is based are often broken or unavailable.

Even using one standard deviation from the average leaves your Emergency Department in overburden mode one third of the time. This is a failed, outdated approach to assuring appropriate staffing moment by moment.

To cope with a clunky staffing model, your overburdened crew have turned stages of patient flow into silos. They have arranged their work individually in their own individual bulletproof box. And to be of utmost efficiency within their box, they batch their work according to their own worklist of waiting patients. Of course, batching their work adds delay.

Each army of one works at their own best pace on their own pile of work, which, by the way, insulates them from the actual flow of patients into the department. 

And this has been calcified into how to do this better and better, adding more stages / silos. 

This is moving in the wrong direction and quite opposite of the shared flow  of a team working together simultaneously in real time on each patient without any silos at all. 

You will see pockets of great teamwork when you watch the work. And that is the start of building more of it into the department. 

You Aren't Looking for Problems: You are In Search of Excellence!

You see, the point of observation isn't actually, to spot a problem. Problems are everywhere. Anyone can give you a list. Important, yes, very. But not the most important thing.

The point of you getting out there and observing is to find the diamonds in the rough: To discover the legitimate moments of exceptional teamwork and care as it happens, right in front of you. Stuff that doesn't make it into the metric.

And then to come back again and see and gather those diamonds in their various situations often enough to get a clue as to what really makes them.

And as a member of management, to figure out what you can do to help grow those diamonds more often. 

Yes, that's on you.

Always start with the situation of the RN and the Patient. That isn't just by interview, though interviews are crucial. It is also by observation. You have a pair of eyes and ears? You have the superior analytic tools. 

Now you are starting the journey on a solid foundation.