MCHB Conference Webcasts
2005 EMSC Annual Grantee Meeting
April 12-13, 2005
KIMBERLY MIDDLETON: I'm going to do a presentation on the 2002 Emergencies Pediatric Services and Equipment Supplement, which I'll refer to as EPSES. It's a supplement to the National Hospital Medical Ambulatory Care Survey, which I'm calling the HMACS.
This is how I'm going to do, the presentation will go. I'll talk about the background, the EPSES methodology, the results we obtained, and then talk about our next steps.
For background, we started with the guidelines for the American Academy of Pediatrics, which I don't have to go over again. But we only pulled out the ones which we actually had data to show. And I won't read the slide.
Our start point was 1998 Nye Survey. We wanted to see if we could replicate the numbers that were obtained, or if pediatric preparedness had increased since 1998. These are some of the findings that we looked at. In 1998 hospitals without pediatric departments were transferring their critically injured patients. 10% of hospitals were treating these critically ill patients in the PICU, and 7% were even treating these patients in the adult ICU.
Few hospitals had protocols for obtaining peds consults. And equipment for children was more likely to be missing than equipment for adults.
In 2002, HRSA funded our survey, which was to replicate the findings in Nye, like I said. It was added on both 2002 and 2003 to our HMACS survey. Overall, the findings that we had were that the EPSES did not show an increase in estimates. And so what that meant was that there's been no increase in the availability of pediatric services since 1998.
So this was a baseline. It answered our preliminary question, but we wanted to dig around in the data that we had to find out what else we could find.
And so I want to talk about the HMACS methodology, which, again, the EPSES was added to that survey. So the National Ambulatory Medical Care Survey was filled in 1992. It's a yearly survey. We go to emergency departments and outpatient departments. On average we survey 480 hospitals, which yields 35,000 ED visits and 24,000 OPD visits.
The main thing that I wanted to point out is that we have a four‑stage sample design; and that means that we first sample from PSUs, then we sample from hospitals, then EDs and OPDs, and then visits from each one of those. We're actually in the hospital for a four‑week reporting period collecting our data.
The way the EPSES fit into this survey was that it was a short set of questions that were answered by the representative when we went in to pull data for the HMACS. It was self‑administered and we collected the survey at the end of the four‑week reporting period.
Just a little bit about our response rates. We actually had 480 hospitals in sample. Once we went out, we found that 83 of these hospitals did not have an emergency department. 51% refused to participate in our survey, so that gave us an 87% response rate.
These are some concepts that we developed based upon the variables that were available in our survey. The first one is pediatric structure. And we were looking to see if the hospital admits pediatric patients. Then if they do, if they have a separate peds ward or not.
These are two other measures. The first one is pediatric case load. What we were looking to see is how many children were actually seen in the emergency department; and then we took a percentage, which was just the number of children over total visits.
Then for our equipment variables, what we were looking to see is how many emergency departments were totally supplied. And then if they weren't, what percentage of the supplies they actually had. We developed a variable that we called adequately supplied. What we did, we took the median, which was 85%, and those above the median we were saying were adequately supplied and those below we were saying were undersupplied.
And the results that I'm going to show are in terms of these variables. This is a vignette that kind of gives you an overview of where our analysis ended up. And the idea here is if you took a dart and threw it at all of the EDs in the United States , this is what you would come up with. This emergency department sees between 15 and 7500 pediatric patients a year, which is about 20 to 30% of their case load. This hospital is most likely not associated with the pediatric ‑‑ over here.
(Laughter)
They probably do have the emergency medicine attending, but they don't have the pediatric specialties that we're looking for. They have only about 80% of the recommended supplies and only 5% are fully supplied.
Then, again, we have our dart and we're aiming it at pediatric visits. You see that most pediatric visits are hospital EDs that have a rather large pediatric volume. The hospitals more likely to have a separate peds ward and PICU. These EDs do have the pediatric specialties that we're looking for, and they have as many as 91% of the recommended supplies and 20% of these EDs are fully supplied, which is better. It's kind of what we're looking for.
So to talk again about the variable that we developed pediatric structure. We found that 10% of hospitals do not admit pediatric patients. 52% admit but don't have a separate peds ward. And then 38% both admit and have the peds ward.
This slide just gives you a cross between this structure variable that we developed in case load and what we want to show here is that most of the volume is between 10 to 40%. And that there are some hospitals down at the bottom who do not see pediatric patients at all.
This slide is just to show our first vignette where you see that the highest column there is actually the hospitals that don't have a peds ward and have a medium volume in their case load. This is showing to contrast with the next one. Oops. Got carried away. Where you see that over 50% of pediatric visits are to hospitals that have a separate pediatric ward and have a high patient pediatric patient case load.
So to talk a little bit about the services that were available in our survey. 37% of all pediatric visits in our survey were for injury. Only 3% of EDs had a separate pediatric service area. 18% of hospitals had 24‑hour observation unit. 16% had coordinated trauma service and 10% had a PICU.
This pyramid is going to set up the next couple slides. But we wanted to get some idea how many cases we're actually talking about in terms of who is being transferred. So for 2002 there were 28 million pediatric ED visits. Of those, 5% were either admitted or transferred. The rest were treated and released. And then of that 5%, 25% were transferred. So that's the group that we're talking about for the next slide.
So we look to see how many hospitals have written transfer agreements if they weren't able to deal with trauma patients and we see about half of them do, which also means that half of them do not.
I'm going to skip that one. But we also asked the question, either with or without a written transfer agreement how many of you transferred children when you're not equipped to provide care. You'll see that number increases 88%. And our numbers do fall within the competence intervals of the Nye Survey; that's what we're saying before, that our numbers don't show any significant change.
There were three kinds of expertise available in our survey that we asked about, and it was the emergency medicine, pediatric emergency medicine and pediatric attending. And, again, showing the Nye numbers, knowing that our numbers fall within the confidence intervals. As you can see, as was expected, the lowest numbers are for the pediatric emergency attending.
Just a little more information about the doctors that were available. Only 9% had a pediatrician on duty and only 26% of the hospitals have written protocols stating when to call a pediatrician to the ED.
I'm going to skip that one. Pediatric equipment. When we added up all of the numbers, only 10.8% answered yes to all of the equipment. We had 131 pieces of equipment on our survey. And if they left it blank or put that they didn't know, we counted that as no. So the 10.8 is only the ones who definitely said yes. Most hospitals answered yes to having the resuscitation medication chart tape and dose estimation systems. But our lowest number was for airway management. And then we asked specifically do you have these supplies in your ED? And many hospitals wanted to clarify that they don't have them in their ED but they're in the Nick cue or central supply or they can get them from the next door hospital or something. So I added that information.
This is the frequency distribution of supplies. And as you can see it's negatively skewed where most of the hospitals had at least 80% of the supplies that we were requesting.
And this slide was just to cross our structure variable with the supply category. And as you can see those with the separate peds ward are better supplied, which is what we would expect.
Some of the hospitals in our sample that weren't supplied, we dug around for additional information. So for the 6% that did not admit pediatric patients and were under 85% supplied, we looked to see if there were other hospitals within the same county and we did find that all of them had other, had another hospital that either had peds beds or a separate peds ward. And some of the hospitals actually wrote in to say that they have agreements with EMS to just bypass their facility for nearby children's hospitals.
Our next steps are we just received our 2003 data and I'm combining them in order to provide analysis. All of this analysis at this point has been at the hospital level. But with more data we'll be able to take it down to the visit level and talk about medications prescribed, diagnosis and type of injury. That's it. Thank you.
(Applause)