Advance Directives for Older Adults in the Emergency Department: A Systematic Review

Abstract

Background:

It has been more than two decades since the passage of the Patient Self-Determination Act (PSDA) of 1991, an act that requires many medical points of care, including emergency departments (EDs), to provide information to patients about advance directives (ADs).

Objective:

The study objective was to determine the prevalence of ADs among ED patients with a focus on older adults and factors associated with rates of completion.

Methods:

We searched PubMed, Embase, PsycINFO, CINAHL, Web of Science, Medline, and the Cochrane Library. Articles were selected according to the following criteria: (1) population: adult ED patients; (2) outcome measures: quantitative prevalence data pertaining to ADs and factors associated with completion of an AD; (3) location: EDs in the United States; and (4) date: published 1991 or later.

Results:

Of the 258 references retrieved as a result of our search, six studies met inclusion criteria. Rates of patient-reported AD completion ranged from 21% to 53%, while ADs were available to ED personnel for 1% to 44% of patients. Patients aged ≥65 years had ADs 21% to 46% of the time. Sociodemographics (e.g., older age, specific religion, white or African American race, being widowed, or having children) and health status related variables (e.g., poor health, institutionalization, and having a primary care provider) were associated with greater likelihood of having an AD.

Conclusions:

Published rates of AD completion vary widely among patients presenting to U.S. EDs. Patient sociodemographic and health status factors are associated with increased rates of AD completion, though rates are low for all populations.

Introduction

Advance directives (ADs) provide an opportunity to voice a personal opinion about medical decision making concerning one's own care. However the power of this voice is too often underutilized by the patient and medical care team, resulting in care inconsistent with patient values and preferences.¹ An AD is a legal document that specifies what the patient's medical care wishes are and/or with whom the care team should consult if the patient is unable to make those decisions for him or herself. When patients have a written AD they are more likely to receive care in line with their wishes.²

In 1991, President George H. W. Bush signed into law the Patient Self-Determination Act (PSDA), in hopes of increasing rates of AD completion. Emergency departments (EDs) have subsequently been mandated to ask patients upon admission if they (1) have executed an AD, (2) would like a copy placed in the record, and (3) would like information about ADs. Since the PSDA took effect in December of 1991, there has been increased delivery of information concerning ADs to patients of all ages presenting to the ED, as well as a small increase in AD completion rates.³ However, there remain a number of concerns about the effectiveness of the PSDA, with lower than expected AD completion rates, as well as concerns about the increased associated burden on busy EDs and the inability to take the time needed to address this sensitive and complex issue.^3–7

The objective of this study was to perform a systematic review following Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines⁸ to determine the prevalence of ADs among ED patients, with a focus on older adults and factors associated with rates of completion.

Methods

Search terms were compiled with help from experts in library science, emergency medicine, palliative care, and geriatrics. Searches were performed in the following electronic databases: PubMed/MEDLINE, Elsevier/Embase, Wiley/Cochrane Library, Thomson Reuters/Web of Science, EBSCO/CINAHL, EBSCO/PsycINFO, and Ovid/MEDLINE. The complete PubMed/MEDLINE search strategy, upon which the other database searches were also built, is available on request.

The database searches were conducted without language restrictions. Inclusion criteria were as follows: (1) population: adult ED patients; (2) outcome measures: quantitative prevalence data pertaining to ADs and factors associated with completion of an AD; (3) location: EDs in the United States; and (4) date: published 1991 through June 2014.

Two independent reviewers (SMR, JO) screened titles and abstracts for general relevance. In the event of disagreement, studies were blindly sent to a third reviewer (MJM) for a final decision. Full texts of selected articles were retrieved and reviewed for selection according to the full inclusion/exclusion criteria described above. Resolution of disagreements at this stage was reached by consensus. Two individuals (SMR and JO) independently extracted data of relevance to this study. Though we initially sought to perform a meta-analysis, this was not possible due to the small number of eligible articles and the heterogeneity of the samples.

Results

We found 317 articles through database searches and 6 additional articles through citation analysis. Of the 268 articles that remained after duplicates were removed, 241 were excluded because of irrelevance to the topic (see Fig. 1). Strict inclusion/exclusion criteria as outlined above were applied to the full text of 27 articles. Of these, 6^9–14 were eligible for inclusion in our review (see Fig. 1).

FIG. 1.

Flowchart of the process of literature search and extraction of studies meeting the inclusion criteria.

In Table 1 we describe the study setting and methodology. Five of the six studies took place in academic EDs;^9–13 the exception was the Davis study.¹⁴ Lahn et al.¹¹ included two EDs, while the others^9,10,12–14 were single-institution studies. The definition of ADs sought was provided in all^9,11–14 but the Ishihara et al. study.¹⁰ Five of the six studies were prospective,^10–14 while Grudzen et al.⁹ performed a retrospective medical record review.

Table 1.

Methodology and Characteristics of Included Studies

Author, year	Setting/years	Definition of AD	Inclusion	Exclusion	Methodology
Davis, 2012¹⁴	Community Heart Hospital ED, San Antonio, TX / 6 mo period	DNR, POA, legal guardianship	Treated and dispositioned by single physician	Dispositioned by another physician	Prospective interviews with serial patients
Grudzen et al., 2012⁹	Academic tertiary care center ED, NY, NY / Jan 2005 & Jan 2009	“ED EMR had any documentation of known ADs”	Palliative care consultation during stay Jan 2005 or Jan 2009 for patients admitted through the ED	None	Retrospective medical record review
Ishihara et al., 1996¹⁰	Academic tertiary care center ED, Nashville, TN / May – Aug 1994	Not defined	18 years AND serious chronic medical condition OR ≥65 years	1-Too ill to be surveyed 2-Language barriers 3-Refused	Prospective survey of convenience sample during representative shifts
Lahn et al., 2001¹¹	Two urban academic EDs, NY, NY	Living will, health care proxy, DNR	From SNF, inpatient hospice, inpatient chronic care facility to ED	1-From nonresidential center	Prospective survey of consecutive SNF patients
Lloveraet al., 1997¹³	Academic tertiary care ED, Manhasset, NY / Dec 1995 – Feb 1996	“Have you written or signed a statement regarding your medical care if you should become severely ill and unable to communicate?”	≥18 years and able to complete survey independently	1-Declined 2-Unable to complete survey themselves	Prospective survey using convenience sample during representative shifts
Llovera et al., 1999¹²	Academic tertiary care ED, Manhasset, NY / Aug – Oct 1997	‘‘A written or signed statement regarding their medical care if they should become severely ill and unable to communicate’’	≥18 years and able to complete survey independently	1-Unresponsive 2-Incompetent 3-Language barriers 4-Refused to answer survey	Prospective survey using convenience sample during representative shifts

AD, advance directive; DNR, do not resuscitate; ED, emergency department; EMR, electronic medical record, POA, power of attorney; SNF, skilled nursing facility.

In Table 2 we provide sociodemographic characteristics of sample subjects, rates of reported ADs, as well as how often the ADs appeared in the medical record (if reported), and factors associated with increased likelihood of having an AD. A total of 2412 subjects were represented, who were predominately female. Mean or median age ranged from 49 to 79 years. Davis,¹⁴ Ishihara et al.,¹⁰ and Llovera et al.,^12,13 provided patient-reported rates of AD of 21% to 53%. Davis,¹⁴ Grudzen et al.,⁹ Ishihara et al.,¹⁰ and Lahn et al.¹¹ reported rates of available ADs from 1% to 44%. Davis¹⁴ and Ishihara et al.¹⁰ reported both numbers, with the difference being 21% reported with 1% available (Davis) and 53% reported with 23% available (Ishihara).

Table 2.

Demographics and Rates of Reported Advance Directives and Available Advance Directives

Author, year	Subjects	Sex	Age (years)	Race/ ethnicity	AD reported	AD available	Factors associated with AD
Davis, 2012¹⁴	309	M 53% F 47%	60 (median)	White 48% Hispanic 45% Black/AA 6% Other 1%	21%	1%	-
Grudzen et al., 2012⁹	2005: 100 2009: 63	2005: M 41% F 59% 2009: M 44% F 56%	2005: 72 (mean) 2009: 66 (mean)	2005 White 31% Hispanic 33% Black/AA 28% Other 8% 2007 White 35% Hispanic 27% Black/AA 30% Other 8%	-	2005: 23 % (14% DNR, 9% proxy) 2009: 23% (17% DNR, 6% proxy)	-
Ishihara et al., 1996¹⁰	238	M 43% F 57%	67 (median)	White 68% Black/AA 31% Other 1%	53%	23% 21% (≥65)	White (vs. Black)
Lahn et al., 2001¹¹	715	M 35% F 65%	79 (mean)	White 58% Black/AA 27% Hispanic 16%	-	44% (64% DNR, 60% HCP, 12% LW) 46% (≥65)	SNF, female, white, ≥65 years
Llovera et al., 1997¹³	511	M 47.9% F 52.1%	53 (mean)	-	27%	-	Older age, regular physician, perceived poorer health
Llovera et al., 1999 ¹²	476	M 50% F 50%	49 (mean)	-	23%	-	Older age, have children, Jewish, Protestant, no religion, white or African American, has PCP, has specialist physicians, chronic medical problems, life threatening medical problems

AD, advance directive; AA, African American; DNR, do not resuscitate; F, female; HCP, heath care proxy; LW, living will; M, male; PCP, primary care physician; SNF; skilled nursing facility.

Ishihara et al.¹⁰ reported that of the 78% of patients who did not have an AD, 42% did not know what an AD was, and 95% had not discussed ADs with their primary care physicians. Of patients who knew what an AD was, 45% reported their reason for having not completed an AD was procrastination, while 41% reported they did not want one.¹⁰ Llovera et al. (1999)¹² also asked about reasons for not executing an AD: 39% of patients agreed with “never thought about it before;” 39% “wanted to write one up, but hadn't gotten around to it;” while 33% “felt their family or friends would be able to make decisions for them without an AD.”

Lahn et al.¹¹ reported on subjects from 17 skilled nursing facilities (SNFs), each sending at least 10 patients to the ED, and found that the rates of completion varied from 0% to 94%. Among the 5% (39) patients presenting who had indications for intubation or CPR, 56% (22) did not have an AD. Of the 44% (17) with an AD, the treating physicians found that document useful in 82% (14) of the cases.¹¹

Factors associated with an increased likelihood of an AD were explored by Ishihara et al.,¹⁰ Lahn et al.,¹¹ and Llovera et al.^12,13 Ishihara et al.¹⁰ found that 42% of those without an AD did not know, in 1994, what an AD was. In all the studies with the exception of Ishihara et al.¹⁰ older age was associated with an increased likelihood of an AD. Ishihara et al.¹⁰ found the rates of having an AD were 21% for those ≥65 versus 23% for all. Ishihara et al.¹⁰ and Llovera et al. (1999)¹² found a correlation between white race and having an AD, while Llovera et al. (1999)¹² also found a correlation between African American race and having an AD. Both Llovera et al. studies^12,13 found an association between having a primary care physician and higher rates of having an AD. In contrast, Ishihara et al.¹⁰ reported that 95% of patients without ADs stated that they had never spoken with their primary care providers about ADs.

Discussion

The low number of studies that were identified for this review highlights the existing lack of knowledge about the effects of the PSDA on AD completion among older ED patients over the last 23 years. The six studies included differed substantially in their sample characteristics. Grudzen et al.⁹ included ED patients with referrals for palliative care consultations. Lahn et al.¹¹ included patients from SNFs and Ishihara et al.¹⁰ included “high risk” patients. The remaining studies^12–14 reported rates for adult ED patients surveyed who were 18 and older without further breakdown by age. This systematic review includes the rates of AD completion for 2412 patients visiting U.S. EDs from Texas, Tennessee, and New York. The samples largely represented middle- to older-age adults, and included a predominance of women; samples were racially and ethnically diverse.

We were interested in both reported rates of AD completion and AD availability at the time of the ED visit. Reported AD completion rates were varied across studies and were below 30% in all^9,11–14 but the Ishihara et al.¹⁰ study. The Ishihara et al.¹⁰ study included a sicker population, namely patients with serious chronic medical conditions or patients ≥65 years, which may help to explain the higher 53% completion rate.

AD availability rates were reported for Davis,¹⁴ Grudzen et al.,⁹ Ishihara et al.,¹⁰ and Lahn et al.,¹¹ as 1%, 23%, 23%, and 44%, respectively. The Grudzen et al.,⁹ Ishihara et al.,¹⁰ and Lahn et al.¹¹ studies included higher-risk subjects: palliative care patients, patients with serious chronic medical conditions, older patients, and patients from SNFs with a mean age of 79 years. Even among these high-risk samples, ADs were available less than half the time.^9–11

Emergency medicine practitioners need real time access to ADs to assure that patient values and preferences for care are available when needed clinically. Unfortunately, ADs were available 46% of the time, dependent on the population studied. During emergency medical situations there is little time to locate a document that is not readily available.

Factors associated with an increased likelihood of an AD included older age, white race, and having a primary care physician, which may be related to higher health literacy and increased access to health care.

For subjects without an AD, procrastination was cited as the reason by 39% and 45% of subjects in Llovera et al. and Ishihara et al., respectively.^{10, 12} Procrastination, as well as a desire to avoid thinking about death or deterioration in health, were also identified as reasons patients had not completed an AD in two trials of older adults from a senior center that did not meet full inclusion criteria for this systematic review.^15,16

It is notable that there was over a decade between the publication of the first four and final two publications included in this review. During that time there was an effort to include in the Affordable Care Act payment of primary care physicians for counseling on AD.¹⁷ That portion of the law was dropped due to misinterpretation and politicization of the issue.¹⁷ The act of counseling on end-of-life planning remains uncompensated at the time of this publication. A meta-analysis was not conducted due to the limited available data as well as the heterogeneity among the populations studied.

Conclusion

Over two decades after initiation of the PSDA, rates of AD completion among adult patients presenting to the ED are low, and ADs are still rarely accessible to ED staff; this is true even among older, high-risk, and institutionalized patients. Factors such as poor health literacy and procrastination appear to be barriers to AD completion. Given the growing older population and the high avoidable costs and poor quality of life associated with end-of-life care that is inconsistent with the values and preferences of those affected, providers and health care personnel have to do better in ensuring that patients, especially the aged and those with complex comorbidities, are supported in completing ADs, and have them available in the electronic health record.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

References

Perkins

: Controlling death: The false promise of advance directives. Ann Intern Med, 2007; 147:51–57.

Silveira

, Kim

, Langa

: Advance directives and outcomes of surrogate decision making before death. N Engl J Med, 2010; 362:1211–1218.

Silveira

, Wiitala

, Piette

: Advance directive completion by elderly Americans: A decade of change. J Am Geriatr Soc, 2014; 62:706–710.

Teno

, Lynn

, Connors

Jr. , et al.: The illusion of end-of-life resource savings with advance directives. SUPPORT Investigators. Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatment. J Am Geriatr Soc, 1997; 45:513–518.

Glick

, Mackay

, Balasingam

, et al.: Advance directives: Barriers to completion. J N Y State Nurses Assoc, 1998; 29:4–8.

Dimick

: Sorting out advance directives. J AHIMA, 2011; 82:26–30; quiz 31.

Sahm

, Will

, Hommel

: Attitudes towards and barriers to writing advance directives amongst cancer patients, healthy controls, and medical staff. J Med Ethics, 2005; 31:437–440.

Liberati

, Altman

, Tetzlaff

, et al.: The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: Explanation and elaboration. PLoS Med, 2009; 6:e1000100.

Grudzen

, Hwang

, Cohen

, et al.: Characteristics of emergency department patients who receive a palliative care consultation. J Palliat Med, 2012; 15:396–399.

10.

Ishihara

, Wrenn

, Wright

, et al.: Advance directives in the emergency department: Too few, too late. Acad Emerg Med, 1996; 3:50–53.

11.

Lahn

, Friedman

, Bijur

, et al.: Advance directives in skilled nursing facility residents transferred to emergency departments. Acad Emerg Med, 2001; 8:1158–1162.

12.

Llovera

, Ward

, Ryan

, et al.: Why don't emergency department patients have advance directives?. Acad Emerg Med, 1999; 6:1054–1060.

13.

Llovera

, Mandel

, Ryan

, et al.: Are emergency department patients thinking about advance directives?. Acad Emerg Med, 1997; 4:976–980.

14.

Davis

: Emergency department visits: We are not prepared. Am J Emerg Med, 2012; 30:1364–1370.

15.

Hamel

, Guse

, Hawranik

, et al.: Advance directives and community-dwelling older adults. West J Nurs Res, 2002; 24:143–158.

16.

Sachs

, Stocking

, Miles

: Empowerment of the older patient? A randomized, controlled trial to increase discussion and use of advance directives. J Am Geriatr Soc, 1992; 40:269–273.

17.

Piemonte

, Hermer

: Avoiding a "death panel" redux. Hastings Cent Rep, 2013; 43:20–28.