• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
J Am Geriatr Soc. Author manuscript; available in PMC May 4, 2009.
Published in final edited form as:
PMCID: PMC2676906
NIHMSID: NIHMS97892

A FRAILTY INDEX PREDICTS SOME BUT NOT ALL ADVERSE OUTCOMES IN OLDER ADULTS DISCHARGED FROM THE EMERGENCY DEPARTMENT

Abstract

Object

The goal of this study was to determine whether frail older adults, based on a deficit accumulation index (DAI), are at increased risk of adverse outcomes following discharge from the emergency department (ED).

Design

Secondary analysis of data from the Medicare Current Beneficiary Survey

Participants

1851 community-dwelling, Medicare fee-for-service enrollees, ≥ 65 years old who were discharged from the ED between January 2000 and September 2002.

Measurements

Primary dependent variable was time to first adverse outcome, defined as repeat outpatient ED visit, hospital admission, nursing home admission or death within 30 days of the index ED visit.

Results

Time to first adverse outcome was shortest among individuals with the highest number of accumulated deficits. The most frail participants were at greater risk of adverse outcomes following ED discharge compared to those who were least frail (HR 1.44, CI 1.06, 1.96). The most frail individuals were also at higher risk of serious adverse outcomes defined as hospitalization, nursing home admission or death (HR 1.98, CI 1.29, 3.05). In contrast, no association was detected between degree of frailty and repeat outpatient ED visits within 30 days (HR 1.06, CI 0.73, 1.54).

Conclusion

The DAI as a construct of frailty was a robust predictor of serious adverse outcomes in the first 30 days after ED discharge. Frailty was not found to be a major determinant of repeat outpatient ED visits; therefore, additional study is needed to investigate this particular type of health service use among older adults.

Keywords: emergency department, health services utilization, deficit accumulation index, frailty

INTRODUCTION

Frail older adults are increasingly recognized as a population at significant risk for adverse outcomes such as functional decline, hospitalization and death.1 Although considerable controversy exists about how best to identify frail individuals, one promising method is a deficit accumulation count.2 This concept of frailty posits that the “count of what is wrong with you” is more important in contributing to vulnerability than specific conditions or physical limitations.3 Originally based on the components of a comprehensive geriatric assessment, the deficit accumulation index has now been operationalized using a variety of data sources and been shown to be robust in its prediction of poor health outcomes including institutionalization and death.2,46 Although the relationship between frailty as defined by the accumulation of deficits and mortality has been extensively examined, relatively little is known about the relationship between frailty and adverse outcomes in populations identified in specific clinical settings.

One important location for this type of research is the Emergency Department. ED use by older adults has risen dramatically over the past decade.7 In addition to being time- and resource-intensive, emergency care of older adults is frequently complicated by underlying chronic medical conditions and unmet social and physical needs.8 However, despite the complexities of their care, between one-half and two-thirds of older patients are discharged from the ED after a diagnosis and treatment plan have been formulated.9 Older adults who are discharged from the ED are at substantial risk for subsequent adverse outcomes including repeat health service use and death.1015 Understanding whether frailty confers additional risk of adverse outcomes in this population could have important implications for risk screening as well as for ED resource planning and policy.

Because patients discharged from the ED frequently experience repeat ED visits that do not result in admission to the hospital as well as more serious events such as hospitalization, this population presents an opportunity for examining the association between frailty and different types of health service utilization. Hospitalization and nursing home admission are often due to serious illness or injury and are often considered “non-discretionary” types of health services. In contrast, ED visits that do not result in hospitalization (hereafter referred to as “outpatient ED visits”) are often for less acute conditions and, therefore, they may be more sensitive to other factors such as an individuals’ living situation, economic status or access to health care. The frailty index is a comprehensive measure of the whole health of individuals, however, it does not account for non-health related factors that may play a prominent role in outpatient service use. Thus, it is unknown whether the relationship between frailty and repeat outpatient ED visits may differ from other types of health service use.

This study was undertaken to determine whether frail older adults, based on a deficit accumulation index, are at increased risk of adverse outcomes following discharge from the ED. Specifically, our goal was to examine the association between frailty and (1) any adverse outcome, (2) outpatient ED visits, and (3) more serious events defined as hospitalization, nursing home admission or death.

METHODS

Design and Setting

This study was a secondary analysis of data from the Medicare Current Beneficiary Survey (MCBS) Cost and Use files and linked Medicare claims. The MCBS is a continuous survey of a nationally representative sample of Medicare beneficiaries drawn from the Centers for Medicare and Medicaid Services (CMS) enrollment file.16 The sample is stratified by age (with over-sampling of persons 85 and older) and drawn within zip code clusters designated as primary sampling units.16 Beneficiaries (or their proxies) are interviewed in person three times a year. The results from the survey are then combined with Medicare administrative claims data to provide additional information such as health care utilization event dates.16 MCBS survey and associated claims data were obtained from CMS following approval of data use agreement #17470. Approval for the study was also obtained from the Institutional Review Board of Duke University Medical Center.

Study sample

The study sample was restricted to community-dwelling subjects who were age-entitled to Medicare (i.e. 65 and over), not enrolled in a Medicare HMO and had at least one outpatient ED visit between January 2000 and September 2002. Subjects enrolled in a Medicare HMO plan were excluded because they did not have fee-for-service bills generated and therefore it was not possible to determine the dates of their health service use. Residents of long-term care facilities were excluded because a different data collection instrument was used for these individuals. Outpatient ED visits were based on Medicare claims.17 MCBS operates on a 4-year rotating panel design; therefore, subjects enter and leave the survey each year. To be included in this sample, participants had to be in the survey the year preceding their ED visit, so that data for predictor variables would be available, and for at least 90 days following their ED visit, so that outcome variables would be available. The cohort was originally constructed for a different study that used 90 day outcomes.18

Measurements

Deficit Accumulation Index

Frailty was defined by a Deficit Accumulation Index (DAI) which was constructed based on established methods for creating similar indices in other databases described by Rockwood and colleagues.2,4,6 The DAI was calculated based on the proportion of deficits present in an individual out of a possible 44 items considered (Table 1). Each health condition or impairment constituted one deficit on the DAI. Health conditions were defined as affirmative responses to items such as “Ever told you had diabetes?”. For items related to mobility and activities of daily living, participants were considered impaired (score=1) if they reported any difficulty with the task or inability to independently perform the task. For other items measured with an ordinal scale, two clinicians agreed on the response level that constituted an impairment. A higher proportion of deficits indicated a higher degree of frailty. The sample was then divided into 4 groups using the quartiles of the DAI measure. Reference cell coding was used to estimate the hazard for each quartile of increasing DAI, using participants who were least frail as the reference group.

Table 1
Selected characteristics of Medicare beneficiaries with at least one ED visit between January 2000 and September 2002, N=1851

Adverse Outcomes

The primary dependent variable was time to first adverse outcome, defined as repeat outpatient ED visit, hospital admission, nursing home admission or death, within 30 days of the index ED visit. The time frame of 30 days was chosen to reflect the time period when adverse events were most frequent18 and most likely to be related to the index ED visit. To explore whether frailty is equally predictive of different types of events, for additional analyses adverse events were further divided into the following distinct categories: (1) outpatient ED visits (ED visits that do not result in hospitalization) and (2) hospitalization, nursing home admission or death. Hospital admission, nursing home admission and death were considered together because all are serious outcomes and there were insufficient numbers of events to permit separate analyses of nursing home admissions and deaths.

Covariates

Models were adjusted for socio-demographic variables and factors that have been previously shown to be important independent predictors of adverse outcomes in this population.18 These included: age (number of years), sex, race (white versus non-white), annual income (above or below the poverty line defined as ≥$10,000 versus <$10,000), living arrangements (alone versus with others), health insurance status (Medicaid versus other), recent outpatient ED visits (last 6 months, yes or no) and recent hospital admissions (last 6 months, yes or no).

Analysis

Sample weights were applied (all proportions presented are weighted) and statistical procedures that accounted for the complex sampling design of the MCBS were utilized for all analyses.19 In constructing the DAI, there was minimal missing data (< 1% missing values for any variable).

Thus, missing data were imputed using the most conservative approach; for any items that were missing, it was assumed the individual did not have the condition or impairment. Following calculation of basic descriptive statistics for all variables, Kaplan-Meier curves were constructed to illustrate event-free survival probabilities among patients in each quartile of the DAI. Cox proportional-hazards regression models were used to examine the relationship between DAI and event hazards (using the lowest quartile of DAI as the reference category). The proportionality assumption was checked by inspecting the graph of the log(−log(survival)) versus log of survival time, which resulted in approximately parallel lines for the 4 quartiles of the DAI. The first model evaluated time to first adverse event using the composite outcome of any outpatient ED visit, hospitalization, nursing home admission or death. The second model evaluated time to first outpatient ED visit. For this analysis, the competing risks of death or hospitalization were accounted for by subtracting inpatient hospitalization days from overall time at risk and censoring participants at time of death. The third model evaluated time to first hospitalization, nursing home admission or death. All models were constructed in three steps. First, only DAI and the outcome of interest were included in the model and crude hazards were estimated. Next, models were adjusted for demographics, living situation and insurance. Finally, recent ED visits and hospitalizations were added to the model to adjust for predisposition to repeat health service use. Results were expressed as hazard ratios (HR) and 95% confidence intervals (CI). Proportional hazards analyses were conducted using STATA® version 10 (StataCorp LP, College Station, TX); all other analyses were conducted using SAS® version 9.1.3 and SAS Enterprise Guide version 4.1 (SAS Institute, Inc., Cary, NC).

RESULTS

Sample characteristics

Of 17,552 subjects identified in MCBS file years 2000–2002 who were community-dwelling, age-entitled to Medicare, and not enrolled in a Medicare HMO, 3,512 (17.7%) had at least one outpatient ED visit between Jan 2000 and Sept 2002. A total of 1661 subjects were excluded from eligibility because 1) they were not in the survey the year preceding their ED visit and, therefore, data to calculate the DAI were not available or 2) they were not in the survey for at least 90 days following their ED visit and, therefore, data for outcome variables were not available. Consequently, a total of 1851 individuals constituted the analysis sample. All 1851 subjects had a DAI calculated and none were lost to follow up. Socio-demographic and clinical characteristics of the sample are displayed in Table 1.

Frailty according to DAI

Out of a possible maximum of 44, the number of deficits ranged from 0–36 (mean 11.3 and median 10). The quartile cutpoints for the DAI scores (expressed as a proportion: deficits present/possible deficits) were: 0.14, 0.23, and 0.36. The 99% limit for number of deficits was 31, representing a sub-maximal proportion of deficits of 0.70.

Frequency and type of adverse outcomes

A total of 394 (20.8% of 1851) beneficiaries discharged from the ED experienced one or more adverse outcomes within 30 days of the index visit (Table 2). The first adverse outcome was outpatient ED visit for 178/394 subjects and hospitalization, nursing home admission or death for 216/394 subjects. Overall, 199 (10.3%) patients returned to the ED but were not admitted, 205 (10.9%) were hospitalized, 27 (1.4%) were admitted to a nursing home, and 38 (2.2%) died within 30 days of ED discharge.

Table 2
Proportion of patients with adverse outcomes within 30 days following discharge from the ED, N=1851

Relationship between DAI and any adverse outcome

The proportion of subjects who experienced any adverse outcome within 30 days of ED discharge increased according to the number of deficits. In the least frail group, 16.2% of subjects experienced an adverse outcome within 30 days compared to 27.4% of those in the most frail group. Time to first adverse outcome was shortest among the group with the highest number of deficits and longest among those with the lowest number of deficits (Figure 1).

Figure 1
Survival curve for first adverse outcome (outpatient ED visit, hospitalization, nursing home admission or death) according to level of frailty, adjusted for demographics, living situation, insurance status and previous health service use (P=0.0036 for ...

Results of the proportional hazards analysis predicting time to first adverse outcome are displayed in Table 3. In unadjusted analyses, patients in the upper two quartiles were at higher risk of adverse outcomes compared to the least frail group HR 1.43 (1.10, 1.87) and HR 1.76 (1.31, 2.36) respectively). Adjusting for demographics, living arrangements, insurance status and recent health service use had modest effects on parameter estimates and confidence intervals.

Table 3
Proportional hazards analysis of first adverse events by deficit accumulation index (DAI),* N=1851

Relationship between DAI and outpatient ED visits

There was little difference in the cumulative proportion of patients who experienced a repeat outpatient ED visit according to the DAI (8.6% in the lowest quartile compared to 9.8% in the highest quartile). Event-free survival curves overlap for most of the immediate post-index ED visit period, although they begin to separate according to DAI at around 20 days (Figure 2a). In unadjusted models, there was a trend toward modestly increased risk of repeat outpatient ED visits among the most frail group (HR 1.31, P=0.14). However, when additional non-health related factors were added to the model, this effect was no longer apparent (Table 3).

Figure 2Figure 2
Figure 2a. Survival curve for repeat outpatient ED visit according to level of frailty, adjusted for demographics, living situation, insurance status and previous health service use (P=0.51 for overall difference between groups).

Relationship between DAI and hospitalization, nursing home admission or death

In contrast, more significant separation according to level of frailty was observed when serious outcomes (hospitalization, nursing home admission or death) were considered (Figure 2b). Higher number of deficits was associated with increased risk of hospitalization, nursing home admission or death in unadjusted and adjusted models (Table 3).

DISCUSSION

These data demonstrate a robust relationship between frailty as defined by a deficit accumulation index and subsequent adverse outcomes among older Medicare beneficiaries discharged from the ED. Risk of hospitalization, nursing home admission or death increased by 10–45% for each quartile of the DAI (i.e. with increasing frailty). In contrast, the risk of repeat outpatient ED visits was not significantly associated with level of frailty after adjustment for demographics, insurance status and previous health service use. These findings suggest that repeat outpatient ED visits are a unique type of health service utilization that may be more significantly influenced by non-health related factors.

Although controversy abounds about how best to define frailty, many agree that its main consequence is an increased risk for adverse health-related outcomes.20 Our findings are consistent with a growing number of studies that have measured frailty in relation to the accumulation of deficits and shown an association with adverse outcomes. Increased level of frailty as measured by deficit accumulation has been previously associated with increased mortality risk using a variety of data sources and patient populations, 2,4,5,2123 but the data on increased risk of institutionalization and higher number of hospital days have been limited to Canadian and Chinese populations respectively.2,4,5,23 The current study’s findings extend previous work in a number of ways. First, these data demonstrate that frailty is associated with near-term (30-day) risk of adverse outcomes, specifically hospitalization, nursing home admission and death, in a US Medicare population. Second, these data indicate that frailty is predictive of subsequent adverse outcomes among subjects who were selected on the basis of an index type of utilization, an outpatient ED visit. Although the data were drawn from a population-based sample, all of the subjects included in this analysis were known health system “users”. Finally, this study’s adjustment for recent ED visits and hospitalizations accounted for the established relationship between previous and future health service utilization and thus increases our confidence with which the relationship between frailty and adverse outcomes can be reported.

This study also provides new insight into the relationship between frailty and an important but under-studied type of outpatient health care use among older adults: outpatient ED visits. Although frail older adults had slightly higher rates of repeat outpatient ED visits, this disparity was accounted for by other characteristics of this population including demographics, insurance status and previous health service use. The fact that such a comprehensive profile of patients’ medical and functional status did not predict repeat outpatient ED use suggests that this particular type of ED utilization may be largely influenced by other factors such as barriers to obtaining follow-up care in other locations. Although there was little separation in event-free survival curves for outpatient ED visits in the period immediately following the index ED visit, the curves begin to separate according to level of frailty around day 20. Patients who return to the ED early after an index visit tend to go back for same problem,12 again suggesting that lack of follow-up care may be playing a role in this type of utilization. As time goes on, frail older adults are at higher risk of developing new conditions and may well utilize more outpatient ED care overall. Without additional clinical data, these issues cannot be addressed definitively and further research in this area is clearly needed. Given that older adults account for more than 8 million outpatient ED visits annually, additional study in this area is of great importance for policy-making and resource-planning.

This study has a number of limitations. First, the DAI was constructed from survey data reported between one and 15 months preceding the index ED visit. Obtaining health status variables from the survey year preceding the index visit was necessary to ensure that these represented the beneficiary’s pre-event level of health and function, however it is possible that the health status of some beneficiaries may have changed during this time. Second, because this sample was restricted to individuals who were aged 65 or over, community-dwelling and Medicare fee-for-service enrollees, this study does not provide information about younger Medicare beneficiaries, those residing in long-term care facilities and those enrolled in Medicare HMO plans. Third, use of a DAI to define frailty is only one approach and other measures of frailty may produce different results. Finally, further study is needed to determine whether deficit accumulation can be further operationalized and refined into a clinically useful construct.

Despite these limitations, these data provide important insights into the relationship between frailty as measured by the accumulation of deficits and adverse outcomes in a cohort of older Medicare beneficiaries discharged from the ED. Frailty has been defined as an excess vulnerability to stressors, with reduced ability to maintain or regain homeostasis after a destabilizing event.1 These data suggest that an ED visit that does not result in hospitalization, qualifies as a destabilizing event in the life of a frail older adult. Patients with higher numbers of deficits were at increased risk of hospitalization, nursing home admission or death in the 30 days following their discharge from the ED. Repeat outpatient ED visits were not predicted by level of frailty and more study is needed to investigate the determinants of this unique type of health service utilization.

Acknowledgments

Financial Disclosure:

This research was conducted while Dr. Hastings was supported by a VA Health Services Research and Development Career Development Award (CD2 06-019-2) and Drs. Hastings and Whitson were supported by Hartford Geriatrics Health Outcomes Scholar Awards from the AGS Foundation for Health in Aging/John A. Hartford Foundation. Dr. Purser was supported by a Mentored Research Scientist Career Development Award (NCMRR/NICHD #5KO1HD04953-02) and Dr. Johnson by a Paul B. Beeson Career Development Award in Aging Research (1K08AG028975-01A1). The authors also gratefully acknowledge support from the Claude D. Pepper Older Americans Independence Center (NIA 1P30 AG028716-01), Duke Aging Center’s Hartford Center of Excellence 2006-0109, Durham VAMC GRECC and the Duke University Junior Faculty Laboratory. The authors thank Tami Swensen, MS of the Research Data Assistance Center for her expert advice on using MCBS data files and Kenneth Schmader, MD for his insights on earlier versions of the manuscript. The views expressed in this article are those of the authors and do not necessarily represent the views of the Department of Veterans Affairs.

Footnotes

Author Contributions

Study concept and design – SNH, JP, KSJ, HEW

Acquisition and/or analysis of data – SNH, RJS

Drafting, revising, and final approval of the article - SNH, JP, RJS, KSJ, HEW

Sponsor’s Role: none

Conflicts of interest: none

References

1. Walston J, Hadley EC, Ferrucci L, Guralnik JM, et al. Research agenda for frailty in older adults: Toward a better understanding of physiology and etiology: Summary from the American Geriatrics Society/National Institute on Aging Research Conference on Frailty in Older Adults. J Am Geriatr Soc. 2006;54:991–1001. [PubMed]
2. Rockwood K, Mitnitski A, Song X, Steen B, Skoog I. Long-term risks of death and institutionalization of elderly people in relation to deficit accumulation at age 70. J Am Geriatr Soc. 2006;54:975–979. [PubMed]
3. Rockwood K. Frailty and Its Definition: A Worthy Challenge. Journal of the American Geriatrics Society. 2005;53(6):1069–1070. [PubMed]
4. Rockwood K, Song X, MacKnight C, et al. A global clinical measure of fitness and frailty in elderly people. CMAJ. 2005;173:489–495. [PMC free article] [PubMed]
5. Rockwood K. Comparing Two Approaches to Measuring Frailty in Elderly People. J Gerontol A Biol Sci Med Sci. 2007 Jul;62(7):738–43. [PubMed]
6. Mitnitski A, Song X, Skoog I, et al. Relative fitness and frailty of elderly men and women in developed countries, in relation to mortality. J Am Geriatr Soc. 2005;53:2184–2189. [PubMed]
7. Cunningham P, May J. Insured Americans drive surge in emergency department visits. Issue Brief #70. October 2003. Center for Studying Health System Change. http://www.hschange.com/CONTENT/613/613.pdf.
8. Wilber ST, Gerson LW, Terrell KM, Carpenter CR, Shah MN, Heard K, Hwang U. Geriatric Emergency Medicine and the 2006 Institute of Medicine Reports from the Committee on the Future of Emergency Care in the US Health System. Acad Emerg Med. 2006;13:1345–1351. [PubMed]
9. 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;39:238–247. [PubMed]
10. Caplan GA, Brown A, Croker WD, Doolan J. Risk of admission within 4 weeks of discharge of elderly patients from the emergency department--the DEED study. Discharge of elderly from emergency department Age Ageing. 1998;27:697–702. [PubMed]
11. Friedmann PD, Jin L, Karrison TG, Hayley DC, Mulliken R, Walter J, Chin MH. Early revisit, hospitalization, or death among older persons discharged from the ED. Am J Emerg Med. 2001 Mar;19(2):125–9. [PubMed]
12. McCusker J, Cardin S, Bellavance F, Belzile E. Return to the emergency department among elders: patterns and predictors. Acad Emerg Med. 2000;7:249–259. [PubMed]
13. Meldon SW, Mion LM, Palmer RM, et al. A Brief Risk-stratification Tool to Predict Repeat Emergency Department Visits and Hospitalizations in Older Patients Discharged from the Emergency Department. Acad Emerg Med. 2003;10:224–232. [PubMed]
14. Hastings SN, Schmader KE, Sloane RJ, Weinberger M, et al. Adverse Health Outcomes After Discharge from the Emergency Department Incidence and Risk Factors in a Veteran Population. J Gen Intern Med. 2007 Nov;22(11):1527–31. [PMC free article] [PubMed]
15. McCusker J, Healey E, Bellavance F, Connolly B. Predictors of repeat emergency department visits by elders. Acad Emerg Med. 1997;4(6):581–8. [PubMed]
16. Adler GS. A profile of the Medicare Current Beneficiary Survey. Health Care Financ Rev. 1994;15:153–163. [PMC free article] [PubMed]
17. How to Identify Emergency Room Services in the Medicare Claims Data, Technical Note: TN-003. [Accessed April 18, 2007]. http://www.resdac.umn.edu/Tools/TBs/TN-003.asp.
18. Hastings SN, Oddone EZ, Fillenbaum GG, Sloane RJ, Schmader KE. Frequency and predictors of adverse health outcomes in older Medicare beneficiaries discharged from the Emergency Department. Medical Care. 2008 in press. [PubMed]
19. Ciol MA, Hoffman JM, Dudgeon BJ, Shumway-Cook A, et al. Understanding the use of weights in the analysis of data from multistage surveys. Arch Physical Med & Rehab. 2006;87(2):299–303. [PubMed]
20. Bergman H, Ferrucci L, Guralnik J, et al. Frailty: An Emerging Research and Clinical Paradigm–Issues and Controversies. J Gerontol A Biol Sci Med Sci. 2007;62:731–737. [PMC free article] [PubMed]
21. Kulminski A, Yashin A, Arbeev K, Akushevich I, Ukraintseva S, Land K, Manton K. Cumulative index of health disorders as an indicator of aging-associated processes in the elderly: results from analyses of the National Long Term Care Survey. Mech Ageing Dev. 2007 Mar;128(3):250–8. Epub 2006 Dec 20. [PMC free article] [PubMed]
22. Goggins WB, Woo J, Sham A, Ho SC. Frailty index as a measure of biological age in a Chinese population. J Gerontol A Biol Sci Med Sci. 2005 Aug;60(8):1046–51. [PubMed]
23. Woo J, Goggins W, Sham A, Ho SC. Public health significance of the frailty index. Disabil Rehabil. 2006 Apr 30;28(8):515–21. [PubMed]
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

  • Cited in Books
    Cited in Books
    PubMed Central articles cited in books
  • MedGen
    MedGen
    Related information in MedGen
  • PubMed
    PubMed
    PubMed citations for these articles

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...