1022
COMMENTARIES
Schmidt, Richardson • RESEARCH AGENDA
factors that cause delays in the
care of patients. The final area of
study is related to output. Many
of the participants saw this as
one of the primary factors lead-
ing to overcrowding. Output fo-
cuses on our inability to move
patients to their new level of care
once a decision has been made.
Specifically, this usually involves
holding patients in the ED who
need inpatient or intensive care
unit care because the hospital is
full. Many in the group saw this
as a recipe for disaster.
The Unraveling Safety Net:
A Research Agenda
t was well past lunch time in ers aware that the nation’s EDs
Atlanta on the last day of the are indeed a part, a big part, of
Society for Academic Emergency that safety net?
I
Medicine annual meeting, yet a
Soon it also became apparent
group of animated and enthusi- that there is no uniform defini-
astic emergency medicine re- tion of the problem of crowding in
searchers, clinicians, and teach- the ED. Without such a defini-
ers sat in a room, postponing tion, it was hard to frame the re-
their meal to brainstorm about search questions. As in many ar-
research ideas and the need to eas of research, such as cardiac
help solve the problems of emer- arrest, we cannot study the pro-
gency department (ED) crowd- cess without standard definitions
ing. Crowding that may be lead- of the relevant process and out-
ing to gaps in care not only for come variables. We have done
the vulnerable patients for whom this with the Utstein criteria in
we may be the only access to cardiac arrest and it is time to do
health care, but for each and this in crowding. In this special
every one of us who may some- issue, Reeder and Garrison offer
day need emergency care as a re- at least one definition of crowd-
sult of an unexpected incident or ing.2 They suggest a number of
serious illness. We present here approaches to measuring patient
a synopsis of some of those ideas, density, acuity levels, and pro-
hoping that they will inspire us vider staffing to determine
to find innovative ways to keep whether crowding is currently a
emergency care available for all problem. Adopting these or simi-
persons when and where they lar definitions would go a long
Having focused on the need
for a definition and a research
agenda surrounding questions of
input, throughput, and output,
the group wanted to think about
the steps necessary to implement
such an agenda. It was clear that
a number of essential activities
should be undertaken immedi-
ately. As previously mentioned,
standardized definitions and cri-
teria must be developed and
agreed upon; regional variation
and differences in individual
hospital characteristics must be
carefully considered during this
process. It was suggested that
many hospitals have been col-
lecting data and tracking various
aspects of this problem; it might
be useful to collate and analyze
whatever administrative data
are available. Like all retrospec-
tive studies, this would suffer
from a lack of complete and con-
sistent data across institutions,
and have the advantage of yield-
ing useful results for relatively
little expense in a short period of
time. Such a review of adminis-
trative data from multiple
sources would undoubtedly help
to refine the standardized defi-
nitions and criteria. However, it
was also clear that ongoing pro-
spective data collection will be
necessary. The most effective ap-
proach would be the develop-
ment of a consortium of emer-
gency researchers and EDs to
address these questions. We
need it.
way toward a uniform standard
Early in the discussion Dr. that would allow us to study the
Sandra Schneider focused us questions facing us.
around the purposes of research
During the morning session,
in this area.1 She suggested con- problems of crowding were
sidering research in a number of viewed as problems of input,
areas: 1) research as the search problems of throughput, or prob-
for the shock story; 2) research lems of output.3 The group found
seeking the root causes of crowd- this a good way to categorize the
ing; 3) research as pure science; areas that need to be considered.
or 4) research into interventions. Input issues in crowding revolve
As she noted, sometimes re- around the increased numbers of
search is done to find the sensa- patients coming to EDs for care.
tional story that will shock the There may be many reasons for
public and policymakers into this. Some areas that need to be
taking note of the problem. Re- explored are: barriers to access-
search can be seen as a means of ing primary care, lack of insur-
advocacy or a means of searching ance, patient motivations such
for the root causes of a problem. as perceived convenience, the ag-
Both may be important. The In- ing of the population, and popu-
stitute of Medicine recently pub- lation growth. Throughput ques-
lished a report on the nation’s tions relate to our efficiency
safety net institutions, yet it while patients are in the ED. In
never considered EDs as safety this area we might consider de-
net institutions. Do we need the lays related to laboratory and ra- therefore recommend developing
shock story to make policymak- diology studies, and any other such a consortium, and call for a