This is the tenth in a series of posts made in preparation for a presentation I will be making for physicians in fellowship training at an upcoming ACCP meeting. In this...
Is your emergency department sick or is it healthy? If you want to know if a patient is sick, you start
with the vital signs. Emergency departments are the same except that the vital signs are data and you have to know what data you need to know in order to assess the health and efficiency of your ER. At my hospital, every morning, I get an email with the data from the previous day. For me, this is just as important as a patient’s morning vitals are to a hospitalist. Some of the key data elements from our hospitals daily report are:
- Number of ER arrivals
- Number and percentage of patients who left without being seen
- Number and percentage of arrivals who were then admitted
- Number and percentage of arrivals who elope
- Number of psychiatry consults and length of stay of psychiatry patients
- Average time from arrival to first provider (physician, NP, or PA)
- Average time from arrival to admission decision
- Average time from admission decision to departure from the ER
- Average length of stay for patients discharged to home
- Overall average length of stay in the ER
- Number of emergency squad arrivals
- Number of patients in ESI 1-5 categories (ESI is the Emergency Severity Index with 1 being the sickest and 5 being the least sick)
- Number of hours of emergency squad diversion
For all of these values, we also have the average daily value over the past 30 days for comparison purposes. It is easy to be overwhelmed by data so you need to know which pieces of data are the most valuable. As the hospital medical director, here are the ones that I scrutinize:
Left without being seen percentage. This is the percentage of patients who show up in your emergency department who leave without being seen by a provider (i.e., a physician, NP, or PA). These are patients who sign in, take a look at how many patients are backed up in the waiting area, and then leave because they don’t want to wait. On the surface, you would think that these are patients with non-acute illnesses, like they think they have a cold or they missed a menstrual period and think they might be pregnant. However, when we have looked at it, a surprising number of these patients are ESI 2 and 3 and really did need to be seen by a provider. If the left-without-being-seen percentage is too high, then your emergency departments healthcare resources are out of alignment with your community’s health care needs. The average ER has about 2.5% of patients who leave without being seen. Although getting to 0% is probably unrealistic for most emergency departments, getting as low as possible is the goal and if your rate is > 2.5%, then you probably have some work to do.
Emergency squad diversion hours. This is the number of hours in a day that emergency squads are “diverted” to other hospitals. Importantly, this does not mean that the hospital is closed and the lay public (and investigative reporters) often confuse this. When an ER goes on diversion, it means that the hospital is not able to easily handle a lot of acutely ill patients and so the squads are diverted to other local hospitals that at that particular time are more able to handle acutely ill patients. In large cities, this works pretty well since hospitals are often just a few miles apart but in a rural area or a one-hospital town, this can result in significant delays in getting patients to a location where they can be managed. Importantly, the hospital will remain open for anyone who does not arrive by emergency squad (for example, walk-in patients) and also generally will remain open for time-limited conditions such as ST elevation myocardial infarctions (STEMIs). There are a lot of reasons why an ER might go on divert: the ER is overwhelmed with patients, multiple patients in cardiac arrest arriving at the hospital ER simultaneously, lack of empty ICU beds to admit critically ill patient to, lack of open regular medical/surgical beds to admit any kind of patients to, unexpected nursing staffing shortages, etc. When the ER goes on divert, not only are you unable to meet your community’s healthcare needs, but since a high percentage of patients arriving by emergency squad end up being admitted, you are turning away potentially lucrative hospital admissions.
Time from arrival to first provider. This tells you how long patients are waiting before they see a doctor (or NP or PA). If this number is too high, then either you don’t have enough providers at certain times of the day, you don’t have enough patient rooms in your ER, or your triage process is not efficient.
Time from arrival to admission decision. This tells you how long it takes your ER providers to decide that a patient needs to be admitted. This metric can be affected by all sorts of things: whether there are enough providers in the ER, how quickly your lab gets blood tests resulted, availability of x-ray and CT scan testing, etc. If this number is too high, then you are going to need to drill down to determine which of the many causes is responsible.
Time from the decision to admit the patient until the patient leaves the ER. This tells you how long it takes to get the patient out of the ER once the ER physician has decided the patient needs to be admitted. Like the last metric, it can be affected by many variables: how quickly a bed in a nursing unit is ready, how efficient your intra-hospital transportation department is, whether your admitting physicians (for example, hospitalists) evaluate patients in the ER before the patient leaves the ER for the nursing unit or whether they see the patient after arrival in the nursing unit, how efficient your admitting department personnel are, etc.
ESI categories. This tells you how sick the patients are that your ER is seeing. In our ER, about three quarters of the patients will have an ESI = 3, followed by ESI 4, ESI 2, and then equally small percentages of ESI 1 & 5. On the other hand, at an ER in a tertiary care hospital or in a trauma hospital, the most common ESI may be 2. If your ER has too many ESI 4’s and 5’s, then it is likely that your community needs more places to for non-acutely ill patients to be seen such as urgent care facilities, primary care physicians who take add-on same day visits, or “minute clinic” facilities such as exist at many pharmacies.
Room turnover per day. You can calculate this from the number of patients seen in the ER (arrivals per day minus the number who left without being seen) divided by the number of patient rooms in the emergency department. In other words, it is how many patients seen on average in each of room in the emergency department. The higher the number, the more efficiently you are moving patients through the ER. But there is a limit and if the number is too high, then it can be a sign that you need more ER beds. Last year, our room turn number was 6.4, in other words, each room in the ER had 6.4 patients every day. That is a pretty high number so we opened several additional rooms in the ER during the busy time of day with a drop in our room turn number to about 5.2 which is more manageable.
Patient satisfaction score. This is a tough one. Almost by definition, the patients don’t want to be in the ER. Compared to patients who come into the hospital for an elective hip replacement who generally leave pretty happy, about the best you can hope for with ER patients is that they don’t leave too unhappy. So it is hard to have as high of a patient satisfaction score for patients seen in the ER compared to those coming in for elective surgery. Nevertheless, if your ER’s patient satisfaction score is low compared to other hospitals emergency rooms, then you’ll need to drill down to find out why. From purely a business standpoint, an unhappy ER patient pays as well as a happy ER patient but the unhappy ER patient is not going to come back to your hospital when he needs a hip replacement.
It’s hard to treat patients without vital signs and it’s hard to do process improvement in your emergency department without data. But equally important, you have to know what data you need and how to interpret that data.
July 25, 2016