Original Contribution

ED visits by older adults for ambulatory care-sensitive and supply-sensitive conditions

Mary W. Carter PhD^a,b,*, Balaji Datti MD, MPH^a,b, Jamie M. Winters MEd^b

^aCenter on Aging, West Virginia University School of Medicine, Morgantown, WV 26506-9127, USA

^bDepartment of Community Medicine, West Virginia University School of Medicine, Morgantown, WV 26506-9190, USA

Received 11 November 2005; revised 15 December 2005; accepted 17 December 2005

Abstract

Objectives: The aim of this study was to examine the effect of advanced age on ED outcomes, including hospitalization for any reason, ambulatory care-sensitive hospitalizations (ACSHs), and supply-sensitive hospitalizations.

Methods: A secondary data analysis of the National Hospital Ambulatory Care Survey was conducted. National estimates of patient visits were obtained using available sampling weights from National Hospital Ambulatory Care Survey, and population estimates were calculated using estimates published by the US Census Bureau.

Results: Older adults made 48 million patient visits to ED between 2000 and 2002. Overall, 20.3% was for an Ambulatory care-sensitive condition, yielding 5 million ACSH, whereas 62% was for a supply-sensitive condition, yielding 9.5 million supply-sensitive hospitalizations. Residents from nursing homes and patients aged 85 years or older were more likely to be hospitalized for any reason, for ACSH, and for supply-sensitive conditions.

Conclusions: Further research is needed to understand how comorbidity contributes to increasing ED and hospital use among older adults.

D 2006

Introduction

Rates of ambulatory care-sensitive hospitalizations (ACSHs) among older adults have increased rapidly over the past decade, from roughly 36.5 ACSH per 1000 adults aged 65 years and older in 1980 to 57.4 ACSH per 1000 older adults in 1998 [1]. These findings hold important

* Corresponding author. Center on Aging, West Virginia University School of Medicine, PO Box 9127, Morgantown, WV 26506-9127, USA. Tel.: +1 304 293 3278; fax: +1 304 293 2700.

E-mail address: [email protected] (M.W. Carter).

implications for both health-care quality and public policy because rates of ACSH are thought to indicate the extent to which populations encounter difficulty in obtaining timely and adequate outpatient ambulatory care services [2] and are used to gauge health-care system performance [3]. Howev- er, because the increase in ACSH is also thought to reflect the aging of the population [4], it remains unclear to what extent increased rates of ACSH among older adults reflect poor access to care vs medical need stemming from a growth in the size and proportion of the population characterized as very old. For example, although adults aged 85 and older were found to have the highest rate of ACSH at 126 per 1000 adults aged 85 years and older [1],

0735-6757/$ – see front matter D 2006 doi:10.1016/j.ajem.2005.12.012

other factors such as geographic location, sex, and race also have been found to affect ACSH rates among older adults [4]. Furthermore, some ambulatory care-sensitive conditions have been identified as highly discretionary or supply sensitive (eg, congestive heart failure [CHF]) [5]. Supply- sensitive hospitalizations have been shown to be influenced by market factors such as per capita supply of Hospital beds and specialty trained physicians [6,7], leading to wide variations in hospitalization rates that cannot be explained by population case-mix differences [8].

Because ED represents an increasingly important source of care for older adults [9], understanding patterns of use and factors associated with hospital admission among older adults represents an ongoing research priority [10]. Moreover, because ED visits by older adults represent an important source of hospital admissions among this population, with roughly 33% to 50% of all ED visits resulting in a hospital stay [11], patterns of ED visits for ambulatory care-sensitive and/or supply-sensi- tive conditions should impart valuable insight into factors contributing to potentially avoidable hospitalizations

Table 1 Sample, national, and per capita estimates of ED use among older adults, 2000-2002

NH indicates nursing home.

^a Sample estimates based on years 2001 and 2002 of NHAMCS data. Population estimates based on 1.5 million nursing home residents per year (Centers for Disease Control and Prevention’s National Center for Health Statistics, 1999 National Nursing Home Survey).

^b Population estimates based on estimated population of adults aged 65 years and older per region (based on aggregation of state-by-state Annual Estimates of the Population By Sex and Age [April 1, 2001, to July 1,2003] [SC-EST2003-02-XX], Population Division, US Census Bureau).

^c Rural population estimate based on US Department of Agriculture Economic Research Service using data from the US Census Bureau, accessed September 17, 2005, from http://www.ers.usda.gov/Briefing/Population/older/.

among older adults. In response, we use a nationwide probability sample of ED visits to explore the effect of advanced age on visit outcomes, including hospitalization for any reason, ACSH, and supply-sensitive hospital- izations. Because a disproportionate number of adults aged 85 years and older reside in skilled nursing facilities [12], and a growing interest in using measures of ACSH to monitor health-care quality among nursing home residents exists [13-16], we examine both community-based and institutional transfers.

Methods

Study design and population

A secondary data analysis of 3 years of data (2000- 2002) from the National Hospital Ambulatory Care Survey was conducted. National Hospital Ambulatory Care Survey has been conducted annually since 1992 by the Centers for Disease Control and Prevention’s National

Variables	Observed count (n)	Estimated no. of occurrences in thousands	Estimated percentage of patient visits	Estimated no. of patient visits per 100 population per year
Characteristics of ED visits
Ambulatory care-sensitive condition	2906	9842	20.3	9.8
Supply-sensitive condition	8935	30329	62.4	30.2
Hospital admission	5234	17620	36.3	17.5
ACSH	1493	5031	10.4	5.0
Supply-sensitive hospitalization	2830	9523	19.6	9.5
Died during ED visit	138	458	0.9	0.5
Patient visit characteristics
Aged z85 y	2879	9877	20.3	74.5
Nonwhite	2558	7571	15.6	63.9
Transfer from NHa	1270	3942	12.2	131.4
Female	8580	29063	59.8	50.1
Urgent visit	4180	14789	30.4	14.7
ED characteristics Located in the Northeastb	3898	10119	20.8	48.0
Located in the Midwestb	3329	12283	25.3	58.0
Located in the Southb	4379	17440	35.9	43.2
Located in the Westb	2787	8762	18.0	41.3
Rurally locatedc	2542	12064	24.8	49.0
For-profit operating status	1320	4493	9.2	4.4
Study sample
Patient visits by adults aged N65 y	14393	48603	100	48.4

Center for Health Statistics. Because NHAMCS is a publicly available data source with no patient or hospital identifiers, exempt status by our institutional review board was granted. The survey uses a 4-stage probability design to achieve a nationally Representative sample of ED and outpatient visits. For the purposes of this study, only data

from the ED files are used. The 4-stage sampling design uses random stratified selection at each stage, beginning with a selection of primary sampling units. The second stage samples short-stay nonfederal/Military hospitals within select primary sampling units. The third stage samples emergency service areas within select hospitals

Table 2 Top 5 most frequent diagnoses among patient visits to the ED by age, nursing home status, and visit type

First Dx % Second Dx % Third Dx % Fourth Dx % Fifth Dx %

All ED patient visits
Aged z 65 y	Chest pain	3.96	CHF	3.13	Pneumonia	2.65	UTI	2.56	Abdominal	2.54
									pain
Aged z 85 y	CHF	4.85	Pneumonia	4.37	UTI	3.51	Chest pain	2.39	Syncope	2.38
NH transfers	Pneumonia	8.77	CHF	6.10	UTI	4.80	GIT bleeding	4.64	Septicemia	4.04
Ambulatory
care-sensitive visits
Aged z 65 y	CHF	15.45	Pneumonia	13.07	UTI	12.64	Bronchitis	6.49	Essential	5.71
									HT
Aged z 85 y	CHF	21.06	Pneumonia	19.01	UTI	15.27	Volume	7.39	Essential	4.50
							depletion		HT
NH transfers	Pneumonia	21.72	UTI	18.65	CHF	17.21	Volume depletion	8.09	Bronchitis	4.80
Supply-sensitive
visits
Aged z 65 y	Chest pain	6.35	CHF	5.01	Abdominal	4.07	Dizziness	2.23	Dyspnea	2.21
					pain
Aged z 85 y	CHF	8.31	Chest pain	4.10	Abdominal	3.75	Volume	2.92	Dyspnea	2.80
					pain		depletion
NH transfers	CHF	7.18	General	5.86	GIT	4.59	Abdominal	4.20	Chest	3.91
			symptom		bleeding		pain		pain
All visits resulting
in hospitalization
Aged z 65 y	Chest pain	7.01	CHF	6.62	Pneumonia	5.33	Stroke	3.79	Syncope	3.31
Aged z 85 y	CHF	8.80	Pneumonia	8.06	UTI	4.13	Stroke	3.77	GIT	3.58
									bleeding
NH transfers	Pneumonia	7.98	CHF	5.48	GIT bleeding	4.32	UTI	4.31	Septicemia	3.84
ACSHs
Aged z 65 y	CHF	23.18	Pneumonia	18.67	UTI	9.14	Coronary	7.83	Bronchitis	7.82
							syndrome
Aged z 85 y	CHF	27.54	Pneumonia	25.23	UTI	12.92	Volume	9.36	Bronchitis	4.43
							depletion
NH transfers	Pneumonia	28.51	CHF	19.84	UTI	15.61	Volume depletion	9.91	Bronchitis	6.07
Supply-sensitive
hospitalizations
Aged z 65 y	Chest pain	12.97	CHF	12.24	GIT	5.93	Bronchitis	5.08	Essential	4.13
					bleeding				HT
Aged z 85 y	CHF	17.38	Bronchitis	7.06	Chest pain	6.35	Volume	5.91	Abdominal	4.43
							depletion		pain
NH transfers	CHF	12.71	GIT	9.66	General	8.02	Volume	6.35	Abdominal	5.10
			bleeding		symptom		depletion		pain

Results based on first diagnosis only. UTI indicates urinary tract infection; GIT, gastrointestinal; HT, hypertension. Percentages represent the proportion of diagnoses observed within each group (row).

Dx, diagnosis.

and the fourth stage samples patient visits within select emergency service areas during 4-week periods. ED parti- cipation rates for all study years is more than 94%, with 376, 391, and 376 hospital EDs participating in 2000, 2001, and 2002, respectively [17-19]. Trained, designated hos- pital staff was responsible for most of the data collection, using standardized collection forms developed by the Census of the Bureau. Field representatives from the Census of the Bureau oversaw data collection and audited data collection records for accuracy and completeness. For most data elements, missing values represented less than 5% of unweighted records [17-19]. Additional details regarding NHAMCS design and methods are available elsewhere [20].

Study sample

The study sample consisted of all patient visits to ED by adults aged 65 years or older between the years 2000 and 2002, resulting in 14 393 patient visits available for study analyses. Because transfers from skilled nursing facilities were not identified in year 2000 of NHAMCS, analyses exploring institution transfers were restricted to years 2001 and 2002, resulting in 10 586 patient visits for related analyses.

Survey content and study measures

The NHAMCS contains a wide variety of information abstracted from each patient visit, including Patient demographic characteristics (age, sex, ethnic background), diagnoses (up to 3 fields stored in the format of International Classification of Diseases, Ninth Revision, Clinical Modification [ICD-9-CM]), reason for visit (up to 3 fields stored in an unique format), urgency of visit, transfer from a skilled nursing facility (years 2001 and 2002 only), disposition at discharge from the ED (admitted to the hospital, death in ED), and Medication orders (up to 6 fields stored in an unique format) as well as hospital operating characteristics such as hospital proprietary status, region of location (Northeast, South, Midwest, West), and location in a metropolitan statistical area or nonmetropolitan statistical area.

Three dependent variables were specified to model outcomes of ED visits by patients aged 65 years and older, including (1) hospitalized for any reason, (2) hospitalized for an ambulatory care-sensitive condition, and (3) hospi-

Statistical analyses

National estimates of patient visits by older adults (aged 65 years and older) to the ED were calculated using available sampling weights in the NHAMCS. Sampling weights in the NHAMCS are designed to adjust for selection bias arising from the complex survey design and nonresponse bias, as well as adjustment of ratios for fixed totals [26]. Yearly per capita Population rates of ED visits by older adults were estimated with data from the US Census Bureau as follows: [(NHAMCS-based national estimate/total base population from the Census)100] [26]. Weighted Pearson v² statistics were calculated to test for (1) differences in visit outcomes by nursing home resident status vs otherwise and (2) differences in visit outcomes by age 85 years and older vs 65 to 84 years old. Weighted multinomial logistic regression was used to model the risk of an ED visit resulting in (1) hospitalization of any type, (2) ACSH, or (3) supply-sensitive hospitalization as a function of patient visit characteristics and ED operating character- istics. Multinomial logistic model results were adjusted for mortality. All SEs of the statistics were weighted to adjust for the complex survey design and nonresponse bias [26].The statistical package STATA 9 (StataCorp, College Station, TX) [27] was used to conduct all analyses.

Results

Between the years 2000 to 2002, roughly 48 million patient visits to ED were made by adults aged 65 years and older across the United States, suggesting an annual rate of approximately 48 visits per 100 older adults (Table 1). Emergency services for ambulatory care- sensitive and supply-sensitive conditions were found to

ACS visit	0.224	0.203	.215 0.230	0.195	.003
Supply-sensitive	0.538	0.638	.000 0.583	0.634	.000

talized for a supply-sensitive condition. Ambulatory care-	ACSH	0.144	0.098	.000 0.142	0.094	.000
sensitive hospitalization was identified after the strategy presented in the Institute of Medicine report, Access to	Supply-sensitive hospitalization	0.226	0.190	.016 0.224	0.189	.003
Health Care in America [21]. Supply-sensitive hospital-

izations were identified using a discretionary classification scheme developed by researchers Ellis and Ash [22,23] for Medicare case-mix reimbursement purposes and which have been used elsewhere for similar study purposes [24,25]. Only the first diagnosis listed was used to classify ACSH or supply-sensitive hospitalizations.

Table 3 Comparing patient visit characteristics using weight-

ed Pearson v²

NH Community P Aged Aged P

z85 y V84 y

visit

Admitted to hospital 0.470

Admitted to ICU 0.050

0.344 .000 0.443 0.342 .000

0.034 .010 0.045 0.040 .251

ACS indicates ambulator care sensitive; NH, nursing home residents;

ACSH, ambulatory care-sensitive hospitalization; CPR, cardiopulmo- nary resuscitation; ICU, intensive care unit.

Urgent status	0.361	0.310	.007 0.340	0.295	.000
Received CPR	0.015	0.005	.009 0.009	0.007	.457
Died	0.017	0.008	.009 0.013	0.008	.076

comprise a sizable portion of those visits, accounting for roughly 40 million ED patient visits by older adults over the 3-year study period. More than one third of all older adult patient visits to the ED resulted in hospital admission, with estimates suggesting roughly 5 million ACSH and nearly 9.5 million hospitalizations identified as supply sensitive occurred.

Table 2 identifies the most frequently occurring diagno- ses by visit type and subpopulation characteristics. Conges- tive heart failure, Bacterial pneumonia, and urinary tract infections consistently ranked among the top 5 most frequently listed diagnoses among all visit types, age groups, and institutional status, except for those patient visits identified as supply sensitive. Likewise, the proportion of all patient visits occurring for CHF, bacterial pneumonia, or urinary tract infection increased among patient visits by adults aged 85 years and older and among older adults transferred from skilled nursing facilities.

Results from the weighted Pearson v² (Table 3) indicated that residents from nursing homes and patients aged 85 years and older were more likely to be admitted to the hospital for any reason, admitted to the hospital with an ambulatory care-sensitive condition, and admitted to the hospital with a supply-sensitive condition. The latter represents an interesting finding as community-based patients and patients aged 84 years and younger were more likely to present in the ED with a condition identified as supply sensitive. Multivariate findings (Table 4) suggest that adults aged 85 years and older experience increased risk for being hospitalized for any reason, as well as

Table 4 Multinomial logistic regression results of hospital admission outcomes among older adult patient visits to the ED

General hospitalizations

RRR (95% CI)

ACSHs

RRR

Supply-sensitive hospitalizations

(95% CI)

RRR

(95% CI)

Patient visit characteristics

Reported estimates adjusted for mortality. Western region served as the omitted reference category. RRR indicates relative risk ratio; CI, confidence intervals; Rx, prescription.

^a Estimate based on 2001-2002 NHMACS only.

* P b .05.

** P b .01.

*** P b .001.

hospitalized for an ambulatory care-sensitive (ACS) condi- tion, whereas transfers from nursing homes were found to experience increased risk for all 3 types of hospitalizations, after controlling for other factors.

Discussion

This study explored patterns of ED use and patient visit outcomes among a nationally representative sample of adults aged 65 years and older. Nearly two thirds of all visits were identified as being either ambulatory care sensitive or supply sensitive, raising questions about the extent to which at least some of the visits potentially may have been avoidable. However, given the strong association of Advancing age, urgent status, and nursing home residency with risk for hospitalization in general and ACSH in particular, additional effort is needed to disentangle the complexity of comorbid conditions and frailty and their effect on managing chronic conditions. To illustrate, the consistency of CHF to appear among the top 3 most frequent ED diagnoses across all subpopulations, age groups, and visit types underscores the dilemma in managing chronic conditions among the very old. For example, although CHF is sensitive to timely access to ambulatory care [28] and broader market supply factors [5], findings presented here with respect to urgent status, advanced age, skilled nursing residency, and increased risk for hospital admission from the ED may reflect the frailty of a population that frequently presents with multiple chronic conditions [29]. Moreover,

Aged N85 y	1.466***	(1.232, 1.746)	1.720***	(1.420, 2.082)	1.155	(0.956, 1.396)
Nonwhite status	0.681**	(0.543, 0.854)	0.769*	(0.605, 0.977)	0.915	(0.745, 1.125)
NH residenta	1.400**	(1.071, 1.828)	2.232***	(1.744, 2.855)	1.417**	(1.103, 1.821)
Female	0.899	(0.779, 1.038)	0.914	(0.777, 1.076)	0.906	(0.766, 1.071)
ED Rx count	0.983	(0.943, 1.024)	1.188***	(1.140, 1.238)	1.027	(0.984, 1.072)
Urgent visit	1.767***	(1.533, 2.037)	1.742***	(1.465, 2.071)	1.983***	(1.708, 2.302)
ED operating
characteristics
Northern region	1.576**	(1.219, 2.038)	1.128	(0.884, 1.441)	1.476***	(1.188, 1.833)
Midwest region	1.178	(0.887, 1.566)	1.048	(0.835, 1.315)	1.355*	(1.049, 1.750)
Southern region	1.263	(0.989, 1.612)	0.939	(0.745, 1.184)	1.152	(0.893, 1.487)
For-profit status	0.886	(0.691, 1.135)	0.821	(0.606, 1.114)	0.941	(0.610, 1.452)
Rurally located	0.642***	(0.512, 0.804)	0.724***	(0.605, 0.867)	0.743*	(0.556, 0.991)

the aging of the population overall must be considered in selecting benchmarks for reducing hospital admissions because, even with Optimal care, patients with CHF will experience exacerbation of symptoms over time [30].

Contrary to study expectations, risk for hospitalization for any reason and ACSH was reduced among patient visits by nonwhite older adults. However, nonwhite older adults appear to disproportionately account for ED visits, representing 15.6% of all ED visits by older adults, although representing only 8.6% of the older adult population. Considered in context, study findings may suggest that nonwhite elders are more likely to seek ED services, perhaps accounting for the higher population ACSH rates reported elsewhere [31-33], but when compared with other patient visits for ACS conditions specifically, nonwhite older adults may be less likely to be admitted, perhaps capturing difficulties in accessing ambulatory care services in the community.

Still, findings presented here underscore the need to improve the monitoring and management of comorbid conditions among older adults, particularly among residents of skilled nursing facilities whose total number of ED visits during the years 2001 and 2002 was twice as large as the actual number of skilled nursing home residents in the United States during the same period. Previous work has highlighted the interface between nursing facilities and hospitals [34], whereas others have recommended reducing rates of hospitalizations [35] among residents of skilled nursing facilities. The findings presented here regarding frequent ED use by residents of nursing facilities for urinary tract infections, bacterial pneumonia, dehydration, CHF, and septicemia support previous calls for targeting conditions leading to potentially avoidable hospitalizations among residents of skilled nursing facilities for care improvement strategies [25]. To the extent that adults aged 85 years and older also experience higher rates of comorbidity and frailty

[36] and similarly high rates of ED use for urinary tract

infections and bacterial pneumonia, efforts to target improved ambulatory care for this population are needed as well.

Although less clear in its implications, the finding that residents of skilled nursing facilities and adults aged 85 years and older are more likely to receive cardiopul- monary resuscitation (CPR) during ED visits raises concern regarding treatment at the end of life. Much recent effort has focused on reducing transfers to hospitals at the end of life and reducing aggressive but futile care [37]. Although this may suggest that the very old are arriving more acutely ill than are younger older adults, as further suggested by the disproportionate percentage of patient visits classified as urgent among those transferred from skilled nursing facilities and among those made by adults aged 85 years and older, further empirical investi- gation was not possible because the number of cases was too few, underscoring the need for future efforts to explore factors that influence the receipt of CPR among very old and frail populations in ED.

Study limitations

Although the current study provides needed insight into the patterns of ED use by various subpopulations of older adults for ambulatory care-sensitive and supply-sensitive conditions, some notable study limitations exist. Because of data constraints, limited information was available to control for population case-mix differences across groups. For example, our finding that nonwhite elders were less likely to be admitted for an ACSH may reflect differences in case-mix acuity between the 2 groups. Similar concerns may exist with respect to the disproportionate use of CPR among the very old and skilled nursing facility popula- tions. Likewise, the lack of available information regarding previous patterns of health-care use, comorbid conditions, length of illness, and so on, among patient visits most likely contributes to the effect size associated with advanced age and skilled nursing home residence, and finally, the study sample of patient visits rather than all older adults prevents generalization of study findings to the population of older adults as a whole. Because only those individuals who actually visit an ED are captured in the current study, little is known about what factors may contribute to the decision to seek ED treatment among older adults, which in turn, may affect study conclusions regarding ED use and outcomes.

Concluding remarks and future directions

Despite noted study limitations, the findings presented here raise important questions about the extent to which rates of ED use by elders for ambulatory care-sensitive and supply-sensitive conditions reflect the aging of the popula- tion and an increase in comorbidity on the one hand vs the effect of poor access to timely and adequate care for chronic conditions on the other hand. Given ongoing efforts to monitor quality of health care by tracking ACSH and supply- sensitive hospitalizations, much effort is need to understand how comorbidity and frailty contribute to ED and hospital use for these conditions. For example, a recent evaluation of clinical care protocols found both failure to recognize neces- sary modifications in treatment recommendations and contra- diction in treatment recommendations for patients presenting with multiple chronic conditions [38]. Consequently, a better understanding of the effects of comorbidity and frailty in advanced age is needed to reduce potentially unnecessary hospitalizations, improve patient care for chronic conditions among the very old, and limit ED and hospital services to those cases requiring immediate acute care.

References

1. Kozak LJ, Hall MJ, Owings MF. Trends in avoidable hospitalizations. Health Aff (Millwood) 2001;20:225 – 32.
2. Billings J, Anderson G, Newman S. Recent findings on preventable hospitalizations. Health Aff (Millwood) 1996;15:239 – 49.
3. Bindman A, Grumbach K, Osmond D, et al. Preventable hospital- izations and access to care. JAMA 1995;274:305 – 11.
4. McCall N, Brody E, Mobley L, et al. Investigation of increasing rates of hospitalization for ambulatory care sensitive hospitalizations among fee-for-service beneficiaries. A final report, CMS Contract No. 500-00-0029, June, 2004.
5. Wennberg JE, Fisher ES, Stukel TA, et al. Use of Medicare Claims data to monitor provider-specific performance among patients with severe chronic illness. Health Aff (Millwood) 2004 Retrieved on September 13, 2005, from:http://content.healthaffairs.org/cgi/reprint/ hlthaff.var.5v1.
6. Fisher ES, Wennberg JE, Stukel TA, et al. Associations among hospital capacity, utilization, and mortality of U.S. Medicare beneficiaries, controlling for sociodemographic factors. Health Serv Res 2000;34:1351 – 62.
7. Wennberg JE, Fisher ES, Skinner JS. Geography and the debate over Medicare reform. Health Aff (Millwood) 2002 Retrieved on February 13, 2002, from http://www.healthaffairs.org/WebExclusives/

Wennberg _Web_Excl_021302.html.

1. Wennberg JE. Unwarranted variations in healthcare delivery: impli- cations for academic medical centers. BMJ 2002;325:961 – 4.
2. McCusker J, Karp I, Cardin S, et al. Determinants of emergency department visits by older adults: a systematic review. Acad Emerg Med 2003;10:1362 – 70.
3. Wilber ST, Gerson LW. A research agenda for Geriatric EMergency medicine. Acad Emerg Med 2003;10:251 – 60.
4. Aminzadeh F, Dalziel WB. Older adults in the emergency department: a systematic review of patterns of use, adverse outcomes and effectiveness of interventions. Ann Emerg Med 2002;39:238 – 47.
5. Murtaugh CM, Kemper P, Spillman BC, et al. The amount, distribution, and timing of lifetime nursing home use. Med Care 1997;35:204 – 18.
6. Carter MW. Factors associated with ambulatory-care sensitive hospitalizations among nursing home residents. J Aging Health 2003;15:295 – 329.
7. Carter MW, Porell FP. vulnerable populations at risk of potentially avoidable hospitalizations: the case of nursing home residents with Alzheimer’s disease. Am J Alzheimers Dis Other Demen 2005;20: 349 – 58.
8. Intrator O, Zinn J, Mor V. Nursing home characteristics and potentially preventable hospitalizations of long-stay residents. J Am Geriatr Soc 2004;52:1730 – 6.
9. Kane RL, Keckhafer G, Flood S, et al. The effect of Evercare on hospital use. J Am Geriatr Soc 2003;51:1427 – 34.
10. NHAMCS micro-data file documentation, 2000. Hyattsville (MD)7 United States Department of Health and Human Services, National Center for Health Statistics; 2002.
11. NHAMCS micro-data file documentation, 2001. Hyattsville (Md)7 United States Department of Health and Human Services, National Center for Health Statistics; 2003.
12. NHAMCS micro-data file documentation, 2002. Hyattsville (Md)7 United States Department of Health and Human Services, National Center for Health Statistics; 2004.
13. McCaig LF, McLemore T. Plan and operation of the National Hospital Ambulatory Medical Care Survey, 1992. Vital and Health Statistics.

Series 1, No. 34. Hyattsville (Md)7 United States Department of Health and Human Services. Nation Center for Health Statistics; 1994.

1. Millman M, editor. access to health care in America. Committee on Monitoring Access to Personal health care services. Institute of Medicine. Washington (DC)7 National Academy Press; 1993.

p. 219-21.

1. Ellis RP, Ash A. Refining the diagnostic cost group model: a proposed modification to the AAPCC for HMO reimbursement: final report. A report to HCFA, cooperative agreement no. 18-C-98526/1-03. Boston: Department of Economics and Health Care Research Unit, School of Medicine, Boston University. 1988.
2. Ellis RP, Ash A. Refinements to the diagnostic cost group model (DCG). Inquiry 1995;32:418 – 29.
3. Porell FP. Geographic variations in hospital discharge rates and discretionary hospital use. J Health Serv Res Policy 1999;4:147 – 53.
4. Carter MW. Factors association with ambulatory care sensitive hospitalizations among nursing home residents. J Aging Health 2003;15:295 – 331.
5. McCaig LF, Burt CW. National hospital ambulatory medical care survey: 2003 emergency department summary. Advanced data, vital and health statistics. No. 358. Hyattsville (Md)7 United States Department of Health and Human Services. Nation Center for Health Statistics; 2005.
6. StataCorp. Stata statistical software: release 9.0. College Station (Tex)7

StataCorp; 2005.

1. Brown AD, Goldacre MJ, Hicks N, et al. Hospitalization for ambulatory care-sensitive conditions: a method for comparative access and quality studies using routinely collected statistics. Can J Public Health 2001;92:155 – 9.
2. Anderson G, Knickman JR. Changing the chronic care system to meet people’s needs. Health Aff (Millwood) 2001;20:146 – 60.
3. Levenson JW, McCarthy EP, Lynn J, et al. The last six months of life for patients with congestive heart failure. J Am Geriatr Soc 2000;48:S101 – 9.
4. Pappas G, Hadden WC, Kozak LJ, et al. Potentially avoidable hospitalizations: inequalities in rates between US socioeconomic groups. Am J Public Health 1997;87:811 – 6.
5. Blustein J, Hanson K, Shea S. Preventable hospitalizations and socioeconomic status. Health Aff (Millwood) 1998;17:177 – 89.
6. Laditka JN, Laditka SB, Mastanduno MP. Hospital utilization for ambulatory care sensitive conditions: health outcome disparities associated with race and ethnicity. Soc Sci Med 2003;57:1429 – 41.
7. Castle NG, Mor V. Hospitalization of nursing home residents: a review of the literature, 1980-1995. Med Care Res Rev 1996;53:123 – 48.
8. Fried TR, Mor V. Frailty and hospitalization of long-term stay nursing home residents. J Am Geriatr Soc 1997;45:265 – 9.
9. Fried LP, Ferrucci L, Darer J, et al. Untangling the concepts of disability, frailty, and comorbidity: implications for improved target- ing and care. J Gerontol A Biol Sci Med Sci 2004;59:255 – 63.
10. Bowman CE, Elford J, Dovey J, et al. Acute hospital admissions from nursing homes: some may be avoidable. Postgrad Med J 2001;77:40 – 2.
11. Boyd CM, Darer J, Boult C, et al. Clinical practice guidelines and quality of care for older patients with multiple comorbid diseases: implications for pay for performance. JAMA 2005;295:716 – 24.