SPECIAL CONTRIBUTION

Quality Indicators for Geriatric Emergency

Care
Kevin M. Terrell, DO, MS, Fredric M. Hustey, MD, Ula Hwang, MD, MPH, Lowell W. Gerson, PhD,
Neil S. Wenger, MD, MPH, and Douglas K. Miller, MD, on behalf of the Society for Academic
Emergency Medicine (SAEM) Geriatric Task Force

Abstract
Objectives: Emergency departments (EDs), similar to other health care environments, are concerned
with improving the quality of patient care. Older patients comprise a large, growing, and particularly
vulnerable subset of ED users. The project objective was to develop ED-specific quality indicators for
older patients to help practitioners identify quality gaps and focus quality improvement efforts.
Methods: The Society for Academic Emergency Medicine (SAEM) Geriatric Task Force, including mem-
bers representing the American College of Emergency Physicians (ACEP), selected three conditions
where there are quality gaps in the care of older patients: cognitive assessment, pain management, and
transitional care in both directions between nursing homes and EDs. For each condition, a content
expert created potential quality indicators based on a systematic review of the literature, supplemented
with expert opinion when necessary. The original candidate quality indicators were modified in response
to evaluation by four groups: the Task Force, the SAEM Geriatric Interest Group, and audiences at the
2007 SAEM Annual Meeting and the 2008 American Geriatrics Society Annual Meeting.
Results: The authors offer 6 quality indicators for cognitive assessment, 6 for pain management, and 11
for transitions between nursing homes and EDs.
Conclusions: These quality indicators will help researchers and clinicians target quality improvement
efforts. The next steps will be to test the feasibility of capturing the quality indicators in existing medical
records and to measure the extent to which each quality indicator is successfully met in current emer-
gency practice.
ACADEMIC EMERGENCY MEDICINE 2009; 16:441–449 ª 2009 by the Society for Academic Emergency
Medicine
Keywords: emergency medical services; emergency service; hospital; geriatrics; health services
for the aged; quality indicators; health care; quality of health care

From the Department of Emergency Medicine (KMT), the Indiana University Center for Aging Research (KMT, DKM), and the
Regenstrief Institute (KMT, DKM), Indiana University School of Medicine, Indianapolis, IN; the Department of Emergency Medi-
cine, Cleveland Clinic Lerner College of Medicine, Case Western Reserve University, Cleveland, OH (FMH); the Department of
Emergency Medicine, Brookdale Department of Geriatrics and Adult Development, Mount Sinai School of Medicine, New York,
NY, the Geriatric Research, Education and Clinical Center, James J. Peters Veterans Administration Medical Center, Bronx, NY
(UH); the Department of Community Health Sciences, Northeastern Ohio Universities College of Medicine, and the Department
of Emergency Medicine, Summa Health System, Akron, OH (LWG); and the Department of Medicine, University of California at
Los Angeles, Los Angeles, CA (NSW).
Received September 3, 2008; revisions received November 10 and November 25, 2008; accepted November 26, 2008.
SAEM Geriatric Task Force members are listed in Appendix A.
Address for correspondence: Dr. Kevin M. Terrell; e-mail: [email protected]. Reprints will not be available.
This project was funded by an award from the American Geriatrics Society (AGS) as part of the Geriatrics for Specialists Initiative,
which is supported by the John A. Hartford Foundation. Drs. Terrell, Hustey, and Hwang are supported by Dennis W. Jahnigen
Career Development Awards, which are funded by the AGS, the John A. Hartford Foundation, and Atlantic Philanthropies.
This SAEM Geriatric Task Force Report was approved by the SAEM Board of Directors in September 2008.
A related commentary appears on page 436.

ª 2009 by the Society for Academic Emergency Medicine ISSN 1069-6563

doi: 10.1111/j.1553-2712.2009.00382.x PII ISSN 1069-6563583 441
442 Terrell et al. • QUALITY GERIATRIC EMERGENCY CARE

T
he quality of health care is increasingly scruti- group of patients. The objective of this project was to
nized.1 Indeed, the principal message of the Insti- develop EM-specific quality indicators for older patients.
tute of Medicine (IOM) reports ‘‘To Err Is We used an approach similar to that used by the ACOVE
Human’’ and ‘‘Crossing the Quality Chasm’’ was that project to develop quality measures dedicated to the
there are substantial problems with the quality of health needs of older patients seeking care in EDs.
care delivered in the United States.2–4 Similar to other
medical specialties, emergency medicine (EM) practice is METHODS
subject to errors and quality concerns.3 The specialty of
EM is characterized by high acuity, high stress, increas- Selection of Conditions to Address with Quality
ing patient volume, and rapid decision-making often with Indicators
incomplete information. These factors, among many oth- The Society for Academic Emergency Medicine (SAEM)
ers, create obstacles to providing high-quality care in and the American College of Emergency Physicians
emergency departments (EDs).5 (ACEP) created the SAEM Geriatric Task Force (Appen-
A prerequisite for assessing (and, where needed, dix A) in part to improve the care delivered to older ED
improving) the quality of emergency medical care is the patients. The Task Force identified common conditions
ability to measure quality of care. An initial step in this for which older adults (defined as those aged 65 years
process is the development of care standards.6 Quality and older for the purposes of this project) seek care in
indicators involve operational definitions to assess the ED and for which there are important quality gaps.
whether care is delivered well or poorly.6 Unlike practice With the understanding that there are many areas for
guidelines that strive to characterize the nuances of best which quality must be improved, the goal was to select
possible care, quality indicators set a minimum standard a small number of important areas to initiate the identi-
for the care expected from clinicians and health sys- fication of quality indicators for the emergency care of
tems.7 Care that does not meet well-constructed quality older adults. The target conditions were chosen based
indicators generally represents low-quality care. on literature review and expert consensus among Task
In addition to the organization of care, quality indica- Force members. The Task Force selected cognitive
tors may be based on processes of care (e.g., timely assessment, palliative care and pain management, and
administration of aspirin to a patient with an acute transitional care in both directions between a nursing
myocardial infarction [MI]) or outcomes of care (e.g., home and an ED as the initial set of conditions faced by
living or dying after an MI).8 Process-defined quality older ED visitors for which quality indicators would be
indicators represent actions of providers, while out- valuable. Palliative care was later removed, as the Cen-
come indicators represent the results of the care pro- ter to Advance Palliative Care has a project with this
cesses plus the effects of many other factors.9 Most focus.
health care quality experts favor process-based quality The Task Force identified a content expert for each
indicators for four reasons. First, processes of care are target condition (cognitive impairment, FMH; pain, UH;
often more efficiently measured. Care processes occur and transitional care, KMT). The content experts cre-
at the time of care delivery, while the interval between ated potential quality indicators using IF-THEN state-
care and outcomes may be long.8,10 Second, process ments, following the ACOVE quality indicator
indicators usually are more sensitive measures of qual- approach.7 The IF statement determines whether a
ity, because a poor outcome does not occur every time patient is eligible for the quality indicator, and THEN
there is a deficiency in a process of care.8 Third, pro- describes the care process that should or should not be
cess-defined quality indicators do not require risk performed. A quality indicator is considered to have
adjustment to the extent that outcome indicators do.11 been satisfied if the medical record indicates that a
Fourth, process-of-care indicators typically are amena- patient is offered or receives the care required by the
ble to direct action by providers, while outcomes quality indicator (e.g., timely administration of aspirin
deficits often are more difficult to address. Thus, pro- for an acute MI). The quality indicator is excluded from
cess-defined quality indicators can drive quality application to the patient if a patient has a documented
improvement efforts by helping direct attention to spe- contraindication to the indicator (e.g., allergy to aspi-
cific, correctable areas that need improvement.10,12,13 rin). The quality indicator is not met if 1) the medical
An ideal set of indicators would be linked to patient record does not indicate that the patient was offered
outcomes through high-quality research;8,12 however, the care required by the indicator, 2) he or she has no
few important care processes have had each aspect documented contraindication, and 3) his or her refusal
rigorously studied.6,14 Quality indicators, therefore, are is not documented in the medical record. The quality
typically developed with the contribution of expert indicators in this project were designed to be used with
opinion.6,12 ED medical records as the data source. The transitional
Quality indicators should target care that has been care quality indicators were designed to use nursing
documented to need improvement. The Assessing Care home medical records as well.
of Vulnerable Elders (ACOVE) investigators found that
vulnerable older persons had substantial deficiencies in Initial Development of Quality Indicators
care, particularly in areas that require specialized geriat- The three content experts conducted systematic reviews
ric care techniques.15 Older patients are particularly vul- for their target condition. They searched for relevant
nerable in the emergency medical system.5 EDs are English language articles in MEDLINE, the Cumulative
major health care providers for seniors; yet, there has Index to Nursing & Allied Health Literature (CINAHL),
been little development of quality measures for this and The Cochrane Library using appropriate subject
ACAD EMERG MED • May 2009, Vol. 16, No. 5 • www.aemj.org 443

headings and text words for each condition. Search RESULTS

terms are provided in Appendix B. For each search, all
titles and abstracts (if available) were reviewed to In the sections that follow, the quality indicators for the
screen for potentially relevant articles. Full texts of three conditions are reported separately. For each con-
potentially relevant articles were examined for possible dition, we provide a brief description of the pertinent
inclusion. Content experts also examined all references literature, the quality indicators, and the rationale for
within relevant articles. After critically reviewing all each indicator or each set of related quality indicators.
applicable articles, each content expert developed a
critical summary of the literature and a preliminary list • Cognitive Assessment
of quality indicators. More than a quarter of older ED patients are cogni-
tively impaired.16–19 Cognitive impairment may be
Serial Revisions of Quality Indicators broadly categorized as delirium or cognitive impair-
Four groups sequentially evaluated the proposed qual- ment without delirium.18,19 Approximately 10% of older
ity indicators: the full SAEM Geriatric Task Force, the ED patients suffer from delirium,17–21 while another
SAEM Geriatric Interest Group, an audience at the 16%–22% have cognitive impairment without delir-
2007 SAEM Annual Meeting, and audience members at ium.17–19 However, cognitive impairment is recognized
a workshop at the 2008 American Geriatrics Society only 28%–38% of the time by EPs.18,19 Identification of
(AGS) annual meeting. The quality indicators were delirium is especially poor (16% to 36% of cases).19–21
modified after each evaluation, based on consideration Delirium is a potentially life-threatening medical
of each group’s responses. First, the literature summa- emergency associated with an increased risk of morbid-
ries and proposed quality indicators were distributed to ity and mortality.22 Patients with unrecognized delirium
the 23 members of the 2006–2007 Task Force. Recipi- who are discharged home from the ED are three times
ents were instructed to critically review the preliminary more likely to die within 3 months than counterparts in
quality indicators and provide feedback to the Task whom delirium is identified by the EP.23 Even in the
Force chair (LWG) or the appropriate content expert, absence of delirium, awareness of a patient’s cognitive
but to avoid replying to all of the Task Force members status is important to ED care, because cognitive
so that all comments would be independent. All feed- impairment without delirium may affect a patient’s abil-
back received by the chair was forwarded to the appro- ity to relay an accurate medical history and carry out
priate content expert. discharge instructions if discharged home from the ED.
The content experts revised the quality indicators in
response to Task Force members’ suggestions. The Quality Indicators for Cognitive Assessment.
revised indicators were distributed to the 30 members Quality Indicator 1: Cognitive Assessment
of the SAEM Geriatric Interest Group. Again, recipients 1. IF an older adult presents to an ED, THEN the
were asked to reply only to the chair or the content ED provider should carry out and document a
expert. The indicators were revised based on the new cognitive assessment (such as an indication of
comments. level of alertness and orientation or an indication
Third, the revised working set of cognitive assessment of abnormal or intact cognitive status) or docu-
and transitional care quality indicators was presented at ment why a cognitive assessment did not occur.
an interactive didactic session at the 2007 SAEM Annual
Meeting. The audience of 36 people included emergency Rationale. Cognitive impairment is common among
physicians (EPs), nurses, and family physicians. The ses- older ED patients,16–19 and cognitive screening is feasi-
sion’s goal was to draw on the expertise of this group to ble16,19 and reasonably accurate24 in the ED setting, yet
refine the quality indicators for use by clinicians, cognitive impairment is identified infrequently by
researchers, educators, and administrators. The session EPs.18–21
began with a background presentation on the nature of Quality
. Indicator 2: Assessment of Patients with Cogni-
quality indicators. The content experts presented their tive Impairment in the ED
sets of proposed quality indicators and the basis for 2. IF an older adult presents to an ED and is found
inclusion of each. A discussion with the audience fol- to have cognitive impairment, THEN an ED care
lowed each presentation. The quality indicators were provider should document whether there has
revised based on these discussions. been an acute change in mental status from base-
The final step was to present the three sets of quality line (or document an attempt to do so).
indicators at a workshop during the 2008 AGS Annual
Meeting. There were 74 persons in the audience, Rationale. Acute change in mental status is a key fea-
including geriatricians, EPs, nurses, social workers, and ture of delirium, a potentially life-threatening medical
nonphysician gerontologists. Similar to the SAEM emergency.22 Patients with delirium generally have
meeting, each set of quality indictors was presented underlying acute medical conditions that require rapid
separately and feedback from the audience was elicited. diagnosis and treatment.25
The group discussion informed modification of quality Quality
. Indicators 3 and 4: ED Care of Patients with
indicators, resulting in the final version presented in Acute Cognitive Impairment Who Are Discharged Home
this article. The process for developing quality indica- IF an older adult presenting to an ED is found to
tors for pain management was the same as that for have cognitive impairment that is a change from base-
cognitive impairment and transitional care, except they line and is discharged home, THEN the ED provider
were not presented at the SAEM meeting. should document the following:
444 Terrell et al. • QUALITY GERIATRIC EMERGENCY CARE

3. Support in the home environment to manage the therapy sessions, and delays in ambulation37 and a
patient’s care. ninefold increase in the risk of delirium.38
4. A plan for medical follow-up. Several organizations, including the AGS, the Ameri-
Rationale. Patients with cognitive impairment may have can Pain Society, the Joint Commission, the IOM, and
difficulty understanding and complying with ED dis- the Agency for Healthcare Research and Quality
charge instructions. Support in the home environment (AHRQ), have issued standards regarding timely,
is likely to be needed to maximize compliance with the tailored, and adequate pain assessment and treat-
ED care plan. Support may also be needed to assist in ment.1,4,39–46 However, pain care remains inconsistent
monitoring the response (or lack thereof) to treatment. and inadequate in many health care settings including
Given the increased mortality risk in patients who are the ED,47–50 and there has been minimal decrease in
discharged home with delirium,21,23 strong consider- overall pain intensity scores nationally.48,51
ation should be given to hospitalizing these patients Advanced age is the strongest predictor of receiving
unless a single precipitant has been identified that can no analgesic for painful conditions in EDs.52 Older
be adequately treated in the home setting with appro- adults are less likely to receive pain medication than
priate social support and medical follow-up.19,25 younger adults with similar conditions involving can-
Quality
. Indicator 5: Detecting Whether Cognitive cer,53 trauma,54 hip fracture,55–57 and orthopedic inju-
Abnormalities Were Previously Recognized ries.58 Indeed, among older ED patients with isolated
long-bone or hip fractures, more than one-third receive
5. IF an older adult presenting to an ED is 1) found to
no analgesia in the ED,52,56,59 and only 57% receive opi-
have an abnormal mental status, 2) has no change
oid medication.56
in mental status from baseline, and 3) is discharged
home, THEN the ED provider should document
Quality Indicators for Pain Management.
whether there has been previous recognition or
Quality Indicators 1–3: Pain Assessment
diagnosis of an abnormal mental status by another
health care provider (or document an unsuccessful 1. IF an older adult presents to the ED, THEN a for-
attempt to determine this). mal assessment for the presence of acute pain
should be documented within 1 hour of arrival to
Rationale. More than 70% of older ED patients with
the ED.
baseline abnormal mental status have no prior history
2. IF an older adult remains in the ED for more than
of impaired cognition.18,19,25 These patients may bene-
6 hours, THEN a second pain assessment should be
fit from medical and social referrals for more formal
documented within 6 hours of arrival in the ED.
cognitive assessment.
3. IF an older adult receives pain treatment while in
Quality
. Indicator 6: ED Care of Patients with Baseline
the ED, THEN a pain reassessment should be docu-
Abnormal Mental Status Who Are Discharged to Home
mented prior to discharge home from the ED.
6. IF an older adult presenting to an ED 1) is found to
Rationale. Several prominent organizations have
have an abnormal mental status that had not been
issued standards regarding pain assessment and treat-
previously recognized or diagnosed by another
ment.1,4,39–46 The standards for pain care include pain
health care provider, 2) has no change in mental
assessments in all patients, the documentation of regu-
status from baseline, and 3) is discharged home,
lar reassessments, and attention to pain symptom man-
THEN a referral for outpatient evaluation of the
agement during discharge planning.60,61 If pain is not
cognitive impairment should be documented.
assessed appropriately, then appropriate treatment
Rationale. More than 70% of ED patients with cogni- cannot be determined.
tive impairment without delirium have no known prior Quality
. Indicator 4: Pain Management
history of cognitive impairment.18,19,26 Although such 4. IF an older adult presents to the ED and has mod-
patients may benefit from medical and social referrals, erate to severe pain (i.e., a numeric rating scale
ED physicians rarely refer these patients for further score of 4 or higher out of 10), THEN pain treat-
evaluation.18 For the small proportion with a potentially ment should be initiated (or the provider should
reversible etiology, early referral for evaluation and document why treatment was not initiated).
treatment can halt progression of cognitive decline and
help preserve cognitive and functional status.27 Even Rationale. Older ED patients are less likely to receive
for patients with Alzheimer’s disease, early diagnosis pain medication than younger adults with similar condi-
and initiation of treatment may prolong functional tions.53–58 More than one-third of older patients with
independence.27 conditions such as isolated long-bone or hip fractures
receive no pain medication during ED visits.52,56,59
• Pain Management Quality
. Indicator 5: Use of Meperidine
Pain, the ‘‘fifth vital sign,’’28,29 is one of the most com-
5. IF an older adult receives analgesic medication while
mon chief complaints in EDs.30,31 Among ED patients
in the ED, THEN meperidine should be avoided.
with pain, nearly three-quarters present with moderate
or severe pain.32,33 Unrelieved acute pain in older adults Rationale. In two studies of older adults who received
is associated with poorer clinical outcomes.34–36 For opioids in an ED, 33% to 72% received meperidine.52,56
example, continued hip fracture pain is associated with For older adults, meperidine is associated with increased
longer hospital stays, missed or shortened physical risk of delirium, fractures, and even death.62–65
ACAD EMERG MED • May 2009, Vol. 16, No. 5 • www.aemj.org 445

Quality
. Indicator 6: Opioid Analgesia and Bowel Rationale. Important written information is typically
Regimen missing in the transfer paperwork when nursing home
residents are transported to EDs.69,70,73 Although these
6. IF an older adult receives an opioid analgesia pre-
four quality indicators do not specify all information
scription upon discharge from the ED, THEN a
needed by emergency providers, the objective was
bowel regimen should also be provided (or the
to list the minimal information that is essential for
provider should document why a bowel regimen
emergency decision-making and treatment in EDs and
was not given).
ambulances and contact information for individuals
Rationale. Constipation is a frequent side effect of that emergency providers may need to speak to
opioid medications and one of the most common fac- urgently.
tors that adversely affect the quality of life for older Quality
. Indicator 5: Medication List
adults.66 Prophylaxis from this effect is particularly
5. IF a nursing home resident is transferred to an ED,
important for the geriatric population.63,67
THEN the nursing home should provide a medica-
tion list in the transfer paperwork.
• Transitional Care
Each year, there are 2.7 million ED visits by residents Rationale. Among nursing home residents who pres-
of nursing homes or other institutions.30 However, ent with at least some transfer documentation, one-
nursing homes and EDs operate independently, provid- quarter arrive without a list of medications.69 Absence
ing care without complete information on the patient’s of a medication list places the patient at risk for drug
condition or medical history or expectations of the interactions and other medication errors.76
other site of care.68 Ten percent of nursing home resi- Quality
. Indicator 6: Tests Requested by Nursing Home
dents are transported to EDs without any documenta- Providers
tion, and essential information typically is missing in
6. IF a nursing home provider requests that specific
the other 90%.69,70 Correspondingly, according to nurs-
tests be performed in the ED, THEN the EP should
ing home providers, nursing home residents often
document performance of the requested tests (or
return from EDs without written documentation.71
document in the medical record why the tests were
With the goal of improving communication between
not performed).
sites of care, several investigators have introduced stan-
dardized transfer forms in interventional studies. The Rationale. Residents who suffer acute illnesses or
transfer forms significantly increased the communica- injuries are generally transported to EDs, in part,
tion of important clinical information;72,73 however, because urgent diagnostic testing is not available in
nursing home staff completed the forms for fewer than most nursing homes.77,78 If emergency providers do not
half of transfers.72,73 Even when the forms were used, perform requested tests, then the patient may be at risk
much of the requested information still was not of repeat ED visits or hospital admissions.68
recorded,72–74 and the exact reason for the ED visit Quality
. Indicator 7: Communication between Nursing
remained missing or unclear in most cases.75 Consid- Home and ED Providers
ered together, these results indicate that improvements
7. IF a nursing home resident will be released from
in quality during transitions of care will involve more
an ED back to the nursing home, THEN the EP
than the simple introduction of a transfer form.
should document communication with a nursing
Patient care in one site affects the care that should
home provider or the primary care or on-call phy-
take place in the other.68 Accordingly, we developed
sician prior to discharge from the ED (or document
transitional care quality indicators that address the care
attempts to do so).
nursing home residents should receive before, during,
and after the ED visit. Rationale. Direct communication between providers
is essential to ensure high-quality transfers of care.79,80
Quality Indicators for Transitional Care. Emergency providers must communicate across care
Quality Indicators 1–4: Critical Data for the Nursing sites to develop a shared care plan and to initiate plan-
Home-to-ED Transfer ning for the next setting before the transfer occurs.68,80
IF a nursing home resident is transferred to an ED, Quality
. Indicators 8 and 9: Critical Data for the ED-to-
THEN the nursing home should provide the following Nursing Home Transfer
written information on the transfer paperwork: IF a nursing home resident is discharged from the
ED back to the nursing home, THEN the ED should
1. Reason for transfer;
provide the following written information in the trans-
2. Code status (i.e., resuscitation status);
fer paperwork:
3. Medication allergies;
4. Contact information for the nursing home, the pri- 8. ED diagnosis;
mary care or on-call physician, and the resident’s 9. Tests performed with results (and tests with pend-
legal health care representative or closest family ing results).
member.
Rationale. According to nursing home providers,
Alternatively, access to this information in an elec- nursing home residents often return from EDs without
tronic medical record would satisfy these quality indica- any written information.71 Although these quality indi-
tors (and the subsequent quality indicators as well). cators do not specify all data needed by nursing home
446 Terrell et al. • QUALITY GERIATRIC EMERGENCY CARE

providers, the goal is to provide important information Although the quality indicators identified through
about the ED visit. this project have face validity, the next phase will be
Quality
. Indicators 10 and 11: Care Provided after ED to determine the feasibility of capturing them in exist-
Visits ing medical records, followed by research to measure
the extent to which each quality indicator is success-
10. IF a nursing home resident is discharged from
fully met in current emergency practice, whether they
the ED back to the nursing home and physician
are associated with important clinical outcomes, and
follow-up is recommended, THEN the patient
whether there is variation across EDs (e.g., commu-
should receive the follow-up (or the medical
nity-based EDs vs. academic EDs). Another future
record should indicate why the follow-up did not
step will be to determine the extent to which quality
occur).
improvement efforts based on each of these indica-
11. IF a nursing home resident is discharged from the
tors are associated with improved clinical outcomes
ED back to the nursing home and the ED provider
in a variety of ED settings. As more information
prescribes or recommends a medication, THEN
becomes available, the quality indicators identified in
the nursing home should administer the medica-
this project may need to be refined, and more will be
tion (or document in the medical record why the
developed in other domains of geriatric emergency
medication was not administered).
care. These findings can then lead to development of
Rationale. EPs discharge nursing home residents with quality-of-care interventions to improve care in areas
the expectation that their care plans will be carried out. of concern.
Uneasiness about what will be accomplished in nursing Notably, no individual physician could be responsible
homes may lead to unnecessary hospitalizations of for performing all of the care processes specified in the
nursing home residents.68 quality indicators.7 Indeed, many of the care processes
might be carried out by a nurse or other ED staff mem-
ber. For example, in most EDs, a triage nurse carries
DISCUSSION out the initial pain assessment of older adults who pres-
ent for care. Similarly, a cognitive assessment could be
This article presents the results of a rigorous process to
performed by a physician, midlevel provider (i.e., physi-
develop quality indicators for cognitive assessment,
cian assistant or nurse practitioner), nurse, or properly
pain management, and transitional care in both direc-
trained social worker.
tions between nursing homes and EDs. We linked sup-
The initial set of unadjudicated indicators was
porting evidence and expert opinion to specify the
developed by individuals with recognized expertise in
minimum care standards for these vulnerable ED
the respective areas, leaving open the possibility of
patients. Quality indicators are most useful when they
bias. To lessen this potential bias, we drew on the
are measures of minimally acceptable care; that is,
expertise of several groups of knowledgeable provid-
when failure to carry out the care specified by a quality
ers to develop and refine the quality indicators. In
indicator would be poor-quality care in nearly all
addition, although the EPs who contributed to these
instances. Such standardized clinical measures are
indicators (including the SAEM audience) were mostly
needed both to improve quality and to facilitate assess-
academic physicians with particular interest in geriat-
ment of progress during quality improvement efforts.1
ric emergency care, the AGS session attendees were
Most currently available measures of quality in EM
a mix of academic and community-based practitio-
focus on the maximum acceptable interval before an
ners. Substantial time was allotted for discussion, and
intervention takes place, such as the time to percutane-
audience members during both group sessions
ous coronary intervention or thrombolysis for patients
(SAEM and AGS meetings) freely offered modifica-
with acute MI.81 Although timeliness is a necessary
tions to the indicators and suggested additional indi-
component of quality care, it is only one aspect of qual-
cators. Importantly, there was general agreement
ity.81 In addition to time-dependent measures of quality,
between the two meetings with respect to the indica-
Lindsay and colleagues82 identified 29 indicators for
tors and what constitutes a minimum standard of
emergency care for asthma, pneumonia, acute MI,
care. Bias could have resulted from a few assertive
chest pain, thromboembolic disease, and ankle ⁄ foot
attendees whose opinions may have influenced others.
trauma. The indicators they developed differ in many
However, the speakers moderated the sessions mind-
respects from the ones we addressed. 1) Their quality
ful of this potential bias and worked to ensure that
indicators are a blend of outcome- and process-based
everyone was provided with the opportunity to com-
measures, with several requiring risk adjustment. We
ment. In addition to the chance to voice opinions, we
prefer process-of-care quality indicators for the reasons
invited audience members to speak to us individually
provided in the introduction. 2) It is unclear whether
after the sessions or to provide written comments.
their clinical indicators specify the ‘‘low bar’’ for emer-
Furthermore, vetting these indicators with several
gency care. Ours were developed to define a minimum
separate groups served to balance any potential bias
standard of care. 3) Their indicators were developed for
from any one group or particularly assertive respon-
application with administrative databases. As advised
dents. Finally, we developed quality indicators for
by the authors of the ACOVE project,7 ours were devel-
three areas in the emergency care of seniors where
oped to use medical records, which generally are sub-
quality improvement is needed; other aspects of geri-
stantially richer data sources than administrative
atric emergency care will require further identification
databases, particularly for older patients.
ACAD EMERG MED • May 2009, Vol. 16, No. 5 • www.aemj.org 447

of quality gaps and development of suitable quality 15. Wenger NS, Solomon DH, Roth CP, et al. The qual-
indicators. ity of medical care provided to vulnerable commu-
nity-dwelling older patients. Ann Intern Med. 2003;
CONCLUSIONS 139:740–7.
16. Gerson LW, Counsell SR, Fontanarosa PB, Smucker
We developed three sets of quality indicators for emer- WD. Case finding for cognitive impairment in
gency medical care that are now ready for feasibility elderly emergency department patients. Ann Emerg
testing. These quality indicators will provide focus and Med. 1994; 23:813–7.
direction to help identify gaps in emergency care of 17. Naughton BJ, Moran MB, Kadah H, Heman-Ackah
older adults. They will help researchers and quality Y, Longano J. Delirium and other cognitive impair-
improvement experts target quality improvement ment in older adults in an emergency department.
efforts. Ann Emerg Med. 1995; 25:751–5.
18. Hustey FM, Meldon SW. The prevalence and docu-
The SAEM Geriatric Task Force included experts in geriatric emer- mentation of impaired mental status in elderly
gency care representing both the Society for Academic Emergency emergency department patients. Ann Emerg Med.
Medicine and the American College of Emergency Physicians. Task
Force members are listed in Appendix A.
2002; 39:248–53.
19. Hustey FM, Meldon SW, Smith MD, Lex CK. The
References effect of mental status screening on the care of
elderly emergency department patients. Ann Emerg
1. Kizer KW. The emerging imperative for health care Med. 2003; 41:678–84.
quality improvement. Acad Emerg Med. 2002; 20. Elie M, Rousseau F, Cole M, Primeau F, McCusker
9:1078–84. J, Bellavance F. Prevalence and detection of delir-
2. Kohn KT, Corrigan JM, Donaldson MS. To Err Is ium in elderly emergency department patients. Can
Human: Building a Safer Health System. Washing- Med Assoc J. 2000; 163:977–81.
ton, DC: National Academy Press, 1999. 21. Lewis LM, Miller DK, Morley JE, Nork MJ, Lasater
3. Burstin H. Crossing the quality chasm in emergency LC. Unrecognized delirium in ED geriatric patients.
medicine. Acad Emerg Med. 2002; 9:1074–7. Am J Emerg Med. 1995; 13:142–5.
4. Institute of Medicine. Crossing the Quality Chasm: 22. Inouye SK. Delirium in older persons. N Engl J
A New Health System for the 21st Century. Wash- Med. 2006; 354:1157–65.
ington, DC: National Academy Press, 2001. 23. Kakuma R, du Fort GG, Arsenault L, et al. Delirium
5. Sanders AB. Quality in emergency medicine: an in older emergency department patients discharged
introduction. Acad Emerg Med. 2002; 9:1064–6. home: effect on survival. J Am Geriatr Soc. 2003;
6. Campbell SM, Braspenning J, Hutchinson A, Mar- 51:443–50.
shall MN. Research methods used in developing 24. Wilber ST, Lofgren SD, Mager TG, Blanda M, Ger-
and applying quality indicators in primary care. son LW. An evaluation of two screening tools for
BMJ. 2003; 326:816–9. cognitive impairment in older emergency depart-
7. Wenger NS, Roth CP, Shekelle P; ACOVE Investiga- ment patients. Acad Emerg Med. 2005; 12:612–6.
tors. Introduction to the assessing care of vulnera- 25. Sanders AB. Missed delirium in older emergency
ble elders-3 quality indicator measurement set. department patients: a quality-of-care problem. Ann
J Am Geriatr Soc. 2007; 55(Suppl 2):S247–252. Emerg Med. 2002; 39:338–41.
8. Brook RH, McGlynn EA, Cleary PD. Quality of 26. Miller DK, Lewis LM, Nork MJ, Morley JE. Con-
health care. Part 2: measuring quality of care. trolled trial of a geriatric case-finding and liaison
N Engl J Med. 1996; 335:966–70. service in an emergency department. J Am Geriatr
9. Zimmerman DR, Karon SL, Arling G, et al. Develop- Soc. 1996; 44:513–20.
ment and testing of nursing home quality indica- 27. Geldmacher DS, Whitehouse PJ. Evaluation of
tors. Health Care Financ Rev. 1995; 16:107–27. dementia. N Engl J Med. 1996; 335:330–6.
10. Higashi T, Shekelle PG, Adams JL, et al. Quality of 28. Veterans Health Administration. Pain as the 5th vital
care is associated with survival in vulnerable older sign toolkit. Available at: http://www1.va.gov/pain_
patients. Ann Intern Med. 2005; 143:274–81. management/index.cfm. Accessed Dec 28, 2006.
11. Graff L, Stevens C, Spaite D, Foody J. Measuring 29. Lanser P, Gesell S. Pain management: the fifth vital
and improving quality in emergency medicine. Acad sign. Healthc Benchmarks. 2001; 8:68–70.
Emerg Med. 2002; 9:1091–107. 30. McCaig LF, Nawar EW. National hospital ambula-
12. Shekelle PG, MacLean CH, Morton SC, Wenger NS. tory medical care survey: 2004 emergency depart-
Assessing care of vulnerable elders: methods for ment summary. Adv Data. 2006; 372:1–29.
developing quality indicators. Ann Intern Med. 31. Burt CW, McCaig LF. Trends in hospital emergency
2001; 135:647–52. department utilization: United States, 1992–99. Vital
13. Wenger NS, Shekelle PG. Assessing care of vulner- Health Stat. 2001; 150:1–34.
able elders: ACOVE project overview. Ann Intern 32. Ducharme J, Barber C. A prospective blinded study
Med. 2001; 135:642–6. on emergency pain assessment and therapy. J
14. U.S. Institute of Medicine Committee for Evaluating Emerg Med. 1995; 13:571–5.
Medical Technologies in Clinical Use. Assessing 33. Guru V, Dubinsky I. The patient vs. caregiver per-
Medical Technologies. Washington, DC: National ception of acute pain in the emergency department.
Academy Press, 1985. J Emerg Med. 2000; 18:7–12.
448 Terrell et al. • QUALITY GERIATRIC EMERGENCY CARE

34. Dworkin RH. Which individuals with acute pain are 52. Jones JS, Johnson K, McNinch M. Age as a risk fac-
most likely to develop a chronic pain syndrome? tor for inadequate emergency department analge-
Pain Forum. 1997; 6:127–36. sia. Am J Emerg Med. 1996; 14:157–60.
35. Katz J, Jackson M, Kavanagh BP, Sandler AN. Acute 53. Cleeland CS, Gonin R, Hatfield AK, et al. Pain and
pain after thoracic surgery predicts long-term post- its treatment in outpatients with metastatic cancer.
thoracotomy pain. Clin J Pain. 1996; 12:50–5. N Engl J Med. 1994; 330:592–6.
36. Desbiens NA, Mueller-Rizner N, Connors AF Jr, 54. Neighbor ML, Honner S, Kohn MA. Factors affect-
Hamel MB, Wenger NS; Help Investigators. Pain in ing emergency department opioid administration to
the oldest-old during hospitalization and up to one severely injured patients. Acad Emerg Med. 2004;
year later. J Am Geriatr Soc. 1997; 45:1167–72. 11:1290–6.
37. Morrison RS, Magaziner J, McLaughlin MA, et al. 55. Ardery G, Herr K, Hannon BJ, Titler MG. Lack of
The impact of post-operative pain on outcomes fol- opioid administration in older hip fracture patients.
lowing hip fracture. Pain. 2003; 103:303–11. Geriatr Nurs. 2003; 24:353–60.
38. Morrison RS, Magaziner J, Gilbert M, et al. Rela- 56. Hwang U, Richardson LD, Sonuyi TO, Morrison RS.
tionship between pain and opioid analgesics on the The effect of emergency department crowding on
development of delirium following hip fracture. J the management of pain in older adults with hip
Gerontol A Biol Sci Med Sci. 2003; 58:76–81. fracture. J Am Geriatr Soc. 2006; 54:270–5.
39. American Pain Society Quality of Care Committee. 57. Feldt KS, Ryden MB, Miles S. Treatment of pain in
Quality improvement guidelines for the treatment cognitively impaired compared with cognitively
of acute pain and cancer pain. JAMA. 1995; intact older patients with hip-fracture. J Am Geriatr
274:1874–80. Soc. 1998; 46:1079–85.
40. AGS Panel on Chronic Pain in Older Persons. The 58. Closs S, Fairtlough HL, Tierney AJ, Currie CT. Pain
management of chronic pain in older persons. J Am in elderly orthopaedic patients. J Clin Nurs. 1993;
Geriatr Soc. 1998; 46:635–51. 2:41–5.
41. Gordon DB, Dahl JL, Miaskowski C, et al. American 59. Selbst SM, Clark M. Analgesic use in the emer-
Pain Society recommendations for improving the gency department. Ann Emerg Med. 1990; 19:1010–
quality of acute and cancer pain management. Arch 3.
Intern Med. 2005; 165:1574–80. 60. Dahl JL. Improving the practice of pain manage-
42. Joint Commission. Setting the Standard for Quality ment. JAMA. 2000; 284:2785.
in Health Care. Available at: http://www.jointcom 61. Joint Commission. JCAHO Comprehensive Accredi-
mission.org/NR/rdonlyres/6C33FEDB-BB50-4CEE- tation Manual for Hospitals: The Official Handbook
950B-A6246DA4911E/0/setting_the_standard.pdf. (updated September 2007). Oakbrook Terrace, IL:
Accessed Jan 11, 2007. Joint Commission on Accreditation of Healthcare
43. Brownfield E. Pain Management Vol 43. Rockville, Organizations, 2008.
MD: Agency for Healthcare Research and Quality, 62. Adunsky A, Levy R, Heim M, Mizrahi E, Arad M.
Contract No. 290-97-0013, 2001. Meperidine analgesia and delirium in aged hip frac-
44. Agency for Health Care Policy and Research. Acute ture patients. Arch Gerontol Geriatr. 2002; 35:253–9.
pain management: operative or medical procedures 63. Terrell KM, Heard K, Miller DK. Prescribing to
and trauma Part 1. Clin Pharm. 1992; 11:309–31. older ED patients. Am J Emerg Med. 2006; 24:468–
45. Agency for Health Care Policy and Research. Acute 78.
pain management: operative or medical procedures 64. Perri M III, Menon AM, Deshpande AD, et al.
and trauma, Part 2. Clin Pharm. 1992; 11:391–414. Adverse outcomes associated with inappropriate
46. Jaycox AK, Carr DB, Payne R, et al. Management of drug use in nursing homes. Ann Pharmacother.
Cancer Pain. Clinical Practice Guideline No. 9. 2005; 39:405–11.
Rockville, MD: Agency for Health Care Policy and 65. Farrar JT. What is clinically meaningful: outcome
Research, 1994. measures in pain clinical trials. Clin J Pain. 2000;
47. Carr DB, Miaskowski C, Dedrick SC, Williams GR. 16:S106–12.
Management of perioperative pain in hospitalized 66. Norton C. Constipation in older patients: effects on
patients: a national survey. J Clin Anesth. 1998; quality of life. Br J Nurs. 2006; 15:188–92.
10:77–85. 67. Williams R, Bosnic N, Duncan AW, Brogan M,
48. Gordon DB, Pellino TA, Miaskowski C, et al. A 10- Cook SF. Prevalence of opioid dispensings and con-
year review of quality improvement monitoring in current gastrointestinal medications in an elderly
pain management: recommendations for standard- population from Ontario, Canada. J Opioid Manag.
ized outcome measures. Pain Manage Nurs. 2002; 2008; 4:193–200.
3:116–30. 68. Terrell KM, Miller DK. Challenges in transitional
49. Wolfe J, Grier HE, Klar N, et al. Symptoms and suf- care between nursing homes and emergency
fering at the end of life in children with cancer. N departments. J Am Med Dir Assoc. 2006; 7:499–505.
Engl J Med. 2000; 342:326–33. 69. Jones JS, Dwyer PR, White LJ, Firman R. Patient
50. Rupp T, Delaney KA. Inadequate analgesia in emer- transfer from nursing home to emergency depart-
gency medicine. Ann Emerg Med. 2004; 43:494–503. ment: outcomes and policy implications. Acad
51. Morrison RS, Meier DE, Fischberg D, et al. Improv- Emerg Med. 1997; 4:908–15.
ing the management of pain in hospitalized adults. 70. Stier PA, Giles BK, Olinger ML, Brizendine EJ, Cor-
Arch Intern Med. 2006; 166:1033–9. dell WH. Do transfer records for extended care
ACAD EMERG MED • May 2009, Vol. 16, No. 5 • www.aemj.org 449

facility patients sent to the emergency department Appendix A

contain essential information? [abstract] Ann Emerg
Med. 2001; 38:S102. 2008–2009 SAEM Geriatrics EM Task Force
71. Davis MN, Toombs Smith S, Tyler S. Improving Jeanne Basior, MD, University of Buffalo
transition and communication between acute care Christopher Carpenter, MD, Washington University
and long-term care: a system for better continuity Michael Cassara, DO, North Shore University Hospital
of care. Ann Long-Term Care. 2005; 13:25–32. Jeffrey Caterino, MD, Ohio State University
72. Terrell KM, Brizendine EJ, Bean WF, et al. An Kathleen Clem, MD, Loma Linda University
extended care facility-to-emergency department James Espinosa, MD, UMDNJ-SOM
transfer form improves communication. Acad Neal Flomenbaum, MD, Weill Cornell Medical College
Emerg Med. 2005; 12:114–8. Adit Ginde, MD, MPH, University of Colorado Denver
73. Cwinn M, Cwinn A, Forster AJ, et al. Prevalence of School of Medicine
information gaps for seniors transferred from nurs- Theresa Gunnarson, MD, Regions Hospital
ing homes to the emergency department [abstract]. Kennon Heard, MD, University of Colorado, School of
Acad Emerg Med. 2006; 13:S32. Medicine
74. Madden C, Garrett J, Busby-Whitehead J. The Teresita Hogan, MD, Resurrection Medical Center
interface between nursing homes and emergency Jason Hughes, MD, University of Iowa
departments: a community effort to improve trans- Fred Hustey, MD, Cleveland Clinic Lerner College of
fer of information. Acad Emerg Med. 1998; 5:1123– Medicine
6. Ula Hwang, MD, MPH, Mount Sinai School of
75. Davis MN, Brumfield VC, Smith ST, Tyler S, Nitsch- Medicine
man J. A One-page nursing home to emergency Sean Kelly, MD, Beth Israel Deaconess Medical Center
room transfer form: what a difference it can make Eve Losman, MD, University of Michigan
during an emergency! Ann Long-Term Care. 2005; Heather Prendergast, MD, University of Illinois at
13:34–8. Chicago
76. Parry C, Coleman EA, Smith JD, Frank J, Kramer Arthur Sanders, MD, University of Arizona
AM. The care transitions intervention: a patient- Manish Shah, MD, MPH, University of Rochester
centered approach to ensuring effective transfers Kirk Stiffler, MD, Northeastern Ohio Universities
between sites of geriatric care. Home Health Care College of Medicine
Serv Q. 2003; 22:1–17. Jeffrey Tabas, MD, San Francisco General Hospital
77. Kayser-Jones JS, Wiener CL, Barbaccia JC. Kevin Terrell, DO, MS, Indiana University
Factors contributing to the hospitalization of Scott Wilber, MD, MPH, Northeastern Ohio
nursing home residents. Gerontologist. 1989; Universities College of Medicine
29:502–10. Robert Woolard, MD, Texas Tech University HSC SOM
78. Buchanan JL, Murkofsky RL, O’Malley AJ, et al. El Paso
Nursing home capabilities and decisions to hospital- Lowell Gerson, PhD—Chair, Northeastern Ohio
ize: a survey of medical directors and directors of Universities College of Medicine
nursing. J Am Geriatr Soc. 2006; 54:458–65.
79. Coleman EA, Boult C. Improving the quality of Appendix B
transitional care for persons with complex care Search terms for cognitive assessment included delir-
needs. J Am Geriatr Soc. 2003; 51:556–7. ium, dementia, altered mental status, confusion, acute
80. Coleman EA, Fox PD. One patient, many places: confusion, cognitive impairment, cognitive dysfunction,
Managing health care transitions, part II: practi- cognitive assessment, cognitive disorders, elderly, and
tioner skills and patient and caregiver preparation. aged. Search terms for pain care included pain care,
Ann Long-Term Care. 2004; 12:34–9. pain management, pain treatment, analgesia, medica-
81. Adams JG, Biros MH. The elusive nature of quality. tions, opiate, and opioid. Search terms for transitional
Acad Emerg Med. 2002; 9:1067–70. care included nursing homes, long-term care, transpor-
82. Lindsay P, Schull M, Bronskill S, Anderson G. The tation of patients, patient transfer, and transitional care.
development of indicators to measure the quality of Additional search terms for all three targeted condi-
clinical care in emergency departments following a tions included emergency medicine; emergency service,
modified-delphi approach. Acad Emerg Med. 2002; hospital; emergency department; and emergency room.
9:1131–9.