Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2017 Aug 28.
Published in final edited form as: J Clin Nurs. 2013 Dec 14;23(5-6):756–765. doi: 10.1111/jocn.12427

Advance Directives Lessen the Decisional Burden of Surrogate Decision Making for the Chronically Critically Ill

Ronald L Hickman Jr 1, Melissa D Pinto 2
PMCID: PMC5573593  NIHMSID: NIHMS599826  PMID: 24330417

Abstract

Aim and objective

To identify the relationships among advance directive status, demographic characteristics, and decisional burden (role stress and depressive symptoms) of surrogate decision makers (SDMs) of patients with chronic critical illness.

Background

Although the prevalence of advance directives among Americans has increased, SDMs are ultimately responsible for complex medical decisions of the chronically critically ill patient. Decisional burden has lasting psychological effects on SDMs. There is insufficient evidence on the influence of advance directives on the decisional burden of surrogate decision makers of patients with chronic critical illness.

Design

The study was a secondary data analysis of cross-sectional data. Data were obtained from 489 surrogate decision makers of chronically critically ill patients at two academic medical centers in Northeast, Ohio, United States between September 2005 and May 2008.

Methods

Data were collected using demographic forms and questionnaires. A single item measure of role stress and the Center for Epidemiological Studies Depression scale were used to capture decisional burden of the SDM. Descriptive statistics, t-tests, chi-square, and path analyses were performed.

Results

Surrogate decision makers who were non-white, with low socioeconomic status, and low education level were less likely to have advance directive documentation for their chronically critically ill patient. The presence of an advance directive mitigates the decisional burden by directly reducing the SDM’s role stress and indirectly lessening the severity of depressive symptoms.

Conclusions

Most SDMs of chronically critically ill patients will not have the benefit of knowing the patient’s preferences for life-sustaining therapies and consequently be at risk for increased decisional burden.

Keywords: advance directives, surrogate decision makers, chronic critically ill, decisional burden, role stress, depressive symptoms, path analysis

INTRODUCTION

Recent advances in critical care therapeutics have brought forth a growing population of American adults with chronic critical illness, expected to reach 600,000 individuals each year by 2020 (Zilberberg, De Wit, Pirone, & Shorr, 2008; Zilberberg, Luippold, & Sulsky, 2008). Chronic critical illness is a life-limiting syndrome that requires continuous mechanical ventilation support for at least 2–4 weeks (Nelson, Cox, Hope, & Carson, 2010), metabolic derangements (Marik, 2010; Schulman & Mechanick, 2012), and transient or permanent states of cognitive impairment (Ehlenbach et al., 2010). Despite the staggering use of healthcare resources for chronically critically ill patients, health related quality of life and survival at one year post-hospitalization remain poor (Chelluri et al., 2004; Cox, Carson, & Lindquist, 2007; Cox et al., 2009; Douglas, Daly, Kelley, O’Toole, & Montenegro, 2007; Douglas, Daly, O’Toole, Kelley, & Montenegro, 2009).

Chronically critically ill patients require dynamic and complex plans of care. Previous studies indicate that a majority of CCI patients will have cognitive impairment that precludes from making healthcare decisions independently (Camhi et al., 2009; Daly et al., 2010; Nelson et al., 2007). Their severe condition and chronic state of cognitive impairment exacerbates the complexity of healthcare decision making and most often necessitates a surrogate decision maker (SDM) to make healthcare decisions on the patient’s behalf (Daly, et al., 2010; Hickman, Daly, & Lee, 2012). For any given hospital admission, SDMs may make healthcare decisions about diagnostic and surgical procedures, use life-sustaining interventions (i.e., artificial nutrition, mechanical ventilation, and hemodialysis) and end-of-life care. Because SDMs often have a close relationship with the patient, the involvement of SDMs in the coordination of a CCI patient’s care is paramount. In that, critical care clinicians will solicit SDMs for guidance on the patient’s treatment preferences to inform care. Regrettably, many SDMs will not have discussions with the patient about his or her preferences towards life-sustaining therapies nor will the patient’s preferences be outlined in a written advance directive (Camhi, et al., 2009). As a result, SDMs are faced with complex healthcare decisions with little or no preparation to make decisions aligned with the patient’s wishes.

BACKGROUND

Recent U.S. healthcare reform legislation, the Patient Protection and Affordable Care and Patient Self-Determination Acts, highlights the importance of delivering healthcare that is aligned with the patient’s preferences and values while maintaining an acceptable health related quality of life. Advance directives (ADs) are a fundamental component of advance care planning that informs patient-centered decision making among families and healthcare providers. These legal documents that provide information on the patient’s preferences for aggressive medical therapies or designate an individual to make healthcare decisions when the patient is seriously ill and cannot make participate in medical decision making. Advance directives have been traditionally viewed as a form of decision support for SDM that are thought to reduce the stress associated with surrogate decision making. Advance directives have a potential benefit not only to the patient, but also to society, because they declare the patient’s preferences, promote shared decision making, enhance the quality of healthcare decisions, and are thought to reduce the use of healthcare resources for patients who do not want aggressive treatment (Barry & Edgman-Levitan, 2012; Kon, 2010; Silveira, Kim, & Langa, 2010).

Despite federal legislation in support of establishing ADs, aging Americans remain reluctant to complete an AD to outline their healthcare preferences for life-sustaining therapies in the event that they become seriously ill and are unable to make their own healthcare decisions (Camhi, et al., 2009). Nor is it usual for CCI patients to discuss their healthcare preferences with SDMs (Camhi, et al., 2009). As a consequence, SDMs are frequently required to make healthcare decisions on the patient’s behalf without guidance and resulting in a heightened state of decisional burden.

Surrogate decision makers report a high negative emotional state associated with having to make a decision under high risk and/or uncertainty, which has been conceptualized as decisional burden. To some extent, decisional burden is the stress or strain associated with being a decision maker and the psychological consequences, such as depression, anxiety, and posttraumatic stress disorder, acquired from making a decision on the behalf of another person (Hickman & Douglas, 2010). This immense decisional burden can have longstanding psychological consequences on individuals who assume the SDM role. Hickman and colleagues (2010) reported significant levels of role stress and depression symptoms among SDMs of CCI patients, which highlight two dimensions of decisional burden.

Among SDMs of patients without an AD, it is speculated that the severity depressive symptoms among SDMs is intensified by their perceptions of the stress associated with making healthcare decisions on the behalf of their critically ill loved one. Factors that influence the appraisal of role stress include inadequate time for deliberation, lack of knowledge of the patient’s preferences for life-sustaining therapies, and increased states of uncertainty (Hickman, et al., 2012; Hickman & Douglas, 2010; O’Connor & Jacobsen, 2007). Role stress and depressive symptoms are two negative psychological consequences that capture the decisional burden of surrogate decision making for CCI patients. To date, few studies (Kramer, Kavanaugh, Trentham-Dietz, Walsh, & Yonker, 2010; Teno, Gruneir, Schwartz, Nanda, & Wetle, 2007; Vig, Starks, Taylor, Hopley, & Fryer-Edwards, 2007) have examined the influence of ADs on mitigating the decisional burden associated with surrogate decision making.

The establishment of ADs has been postulated to reduce the SDM’s decisional uncertainty and alleviate decisional burden associated with making healthcare decisions for patients who lack decisional capacity. There is limited scientific evidence among SDMs of the CCI that establishes direct or indirect effects of ADs on the decisional burden, which has been operationalized as the severity of roles stress and depressive symptoms. Based on the findings of Hickman and colleagues (2010) and a synthesis of literature on advance care planning, and healthcare decision making under uncertainty studies (Kramer, et al., 2010; O’Connor & Jacobsen, 2007; O’Connor, Jacobsen, & Stacey, 2005; O’Connor, Legare, & Stacey, 2005; Teno, et al., 2007; Vig, et al., 2007), we posited that role stress had a direct influence on the SDM’s appraisal of their depressive symptoms and the presence of an AD would blunt the effects of role stress and indirectly attenuate the severity of depressive symptoms among SDM’s of CCI patients with an advance directive.

AIM

The aim of this report is to identify the relationships among AD status, demographic characteristics, and determine the direct and indirect predictive associations among SDMs demographics, AD status on decisional burden (role stress and depressive symptoms) of SDMs of patients with chronic critical illness.

METHODS

Design

This report is a secondary data analysis of a quasi-experimental study among SDMs of CCI patients to examine the effects of structured family meetings as decisional support for surrogate decision makers of CCI patients. In the present study, only baseline data were selected and entered into the statistical analyses to avoid the confounding effects associated with exposure to the experimental condition of the parent investigation. Douglas, Daly, & Lipson (2012), Daly et al., (2010), and Douglas, Daly, O’Toole, & Hickman (2009) describe the methods, procedures, and main outcomes of the parent investigation.

Sample

Surrogate decision makers of CCI patients (n=489) were recruited from five intensive care units at two academic medical centers in Cleveland, OH, USA. The inclusion criteria were adults (>18 years of age), recognized by the medical team as the next-of-kin or legal representative of a decisionally impaired patient (Glasgow coma scale [GCS] <6) who required at least three days of mechanical ventilation in one of five intensive care units. SDMs were excluded if they were: unavailable to meet with the medical team, unable to understand or speak English, or were responsible for making decisions for a critically ill patient that was ventilator dependent prior to admission to the intensive care unit.

Instruments

Demographic characteristics of SDMs

An investigator developed demographic and clinical characteristics form was used for data collection. Data on the SDMs’ age (in years), gender, marital status, race/ethnicity, socioeconomic status, educational level, employment status, caregiver status, relationship to the chronically critically ill patient, and prior experience with a loved one in an intensive care unit.

Decisional Burden

Decisional burden was operationalized by two measures, a single-item measure of role stress and the Center for Epidemiological Studies (CESD) scale for depressive symptoms. Although relational concepts, role stress and depressive symptoms, represent two conceptual dimensions, the acute stress of surrogate decision making and a negative emotional state, of decisional burden previously reported by SDMs of CCI patients.

Role stress

An investigator-derived single-item measure, “How stressful has it been making medical decisions for your loved one?” was used to capture surrogate decision makers’ perception of the psychological stress associated the surrogate decision maker role. This measure was constructed as visual analog scale by two anchors, “not at all stressful” and “very stressful”. Participants were asked to endorse their level of stress associated with the surrogate decision maker role by marking an “X” on a 10 millimeter line to reflect the magnitude of their perceived stress. The scoring for this single-item measure of role stress consisted of finding the center of the “X” mark and then drawing a perpendicular line through the center of the mark and intersects with the horizontal line containing the three anchors. Once the perpendicular line and intersects the horizontal line, the distance from zero to the perpendicular line drawn through the center of the “X” mark was calculated. The distance, calculated in millimeters, was divided by the total length of the scale (10 millimeters), and multiplied by 100 to transform the raw score into a percentage. Higher scores on this measure of role stress indicate increases in the level of psychological stress associated with surrogate decision maker role.

Severity of depressive symptoms

The Center for Epidemiological Studies Depression (CESD) scale was administered to participants as a measure of depressive symptoms (Radloff, 1977). This widely used measure of depressive symptoms has been used in numerous caregiver populations (Choi et al., 2012; Douglas, Daly, O’Toole, & Hickman, 2009; Joling et al., 2012), which includes surrogate decision makers of the CCI (Douglas & Daly, 2005; Douglas, Daly, O’Toole, & Hickman, 2009). This measure consists of 20 items that reflect symptoms associated with major depression. For each item, participants report the amount of time during the previous week that they were affected by a symptom on a four point Likert scale from 0 (rarely or none of the time) to 3 (most or all of the time)”. The CESD total score were calculated by summing the four reversed scored items and the remaining 15 negatively phrased items. Total scores for the CESD range from 0 to 60 and higher scores indicate a greater severity of depressive symptoms (Radloff, 1977).

Ethical approval

The institutional review board of University Hospitals Case Medical Center, Cleveland, OH, USA approved the procedures of this study involving human subjects. Eligible surrogate decision makers were approached by a research assistant (RA) who reviewed the risks and benefits, explained what was expected from subjects, and emphasized the voluntary nature of their participation. If the eligible SDM wished to participate in this study, written informed consent was obtained.

Procedures

Screening and recruitment

Prior to the recruitment of participants, approval was obtained from the institutional review boards at each academic medical center and written informed consent to collect patient data was obtained from the patient’s next-of-kin or legal representative designated to make medical decisions. Patients were screened by research assistants for study eligibility each week day. Once a patient met the eligibility criteria, the research assistant would assess the most current Glasgow Coma Scale (GCS) score to determine the patient’s decisional status. If the patient was sedated, confused, and had an eye score less than three or a motor score less than six on the GCS, the patient was determined to lack decisional capacity. To verify the patient’s impaired decisional capacity, the attending ICU physician was asked to clinically assess the patient’s decisional capacity. Only the next-of-kin or legal representatives of patients who were verified by an ICU physician as lacking decisional capacity were approached for written informed consent to participate in the parent study.

Data collection

After the obtaining of written informed consent from each participant, a face-to-face structured interview was conducted with each participant to administer the demographic and psychosocial questionnaires, which included the single-item measure of role stress and the CESD. These face-to-face structured interviews were approximately 20 minutes in duration. Following the completion of the each structured interview, the research assistant reviewed the patient’s medical records to evaluate documentation of ADs (living will and/or durable power of attorney).

Evaluation of AD documentation

A patient was considered to have an AD if there was either documentation of a living will (LW), durable power of attorney (DPA) for health care, or both in their medical records. If there was no documentation of an LW or DPA in the patient’s chart (either by mention of it in a registered nurse or physician note or by the placement of a copy of the AD in the subject’s chart) then the variable was coded with a score of zero. However, if there was documentation (via registered nurse or physician notes, or copy placed in subject’s chart) that the patient did have a LW or DPA, the variables was coded with a score of one. To assess the differences in demographic and clinical characteristics by AD status, a composite variable of LW and DPA variables was computed by recoding the presence of a LW and/or DPA with a score of one and the absence of LW and DPA with a score of zero.

Data analysis

Statistical analyses were performed using Statistical Package for Social Sciences (SPSS, version 20.0) in conjunction with Analysis of Moment Structures (AMOS, version 20.0). The statistical analyses for this secondary data analysis were conducted in four stages. First, descriptive statistics and measures of central tendency were analyzed to describe the demographic and clinical characteristics of this sample of surrogate decision makers. Second, differences in the mean scores of variables between surrogate decision makers by AD status (No AD vs. AD) were examined using independent samples t tests for continuous dependent variables and chi-square analyses were used for dichotomous dependent variables. Third, bivariate correlations were analyzed to investigate the associations among the study variables.

The final statistical method, a path analysis, was conducted using maximum likelihood estimation to examine the direct and indirect predictive associations among the study variables. The sample size for this study (n =182 with complete data) meets the 10–20 cases per variable criterion needed to conduct a path analysis with sufficient statistical power (Kline, 2011). The initial path analysis model included all variables with bivariate correlation coefficients statistically significant at a level of less than .l0, which enabled the construction of an exploratory path model that could be trimmed for parsimony.

To evaluate the fit of the path analysis model to these data, standard goodness-of-fit indices were examined and used to guide the model trimming procedures. The following goodness-of-fit indices were used to establish the validity of the path model in this sample of CCI patients: chi-square [χ2, p >.05], Tucker Lewis Index [TLI; >.90 acceptable, >.95 excellent], the Comparative Fit Index [CFI; >.90 acceptable, >.95 excellent], and Root Mean Square Error of Approximation [RMSEA; <.08 acceptable, <.05 excellent] (Bentler, 1990; Bentler & Bonnett, 1980; Brown & Cudeck, 1995; Tucker & Lewis, 1973). Standardized indirect and direct regression coefficients and the explained variance for each endogenous variable were calculated. A systematic method of model trimming, which consisted of individually removing statistically insignificant paths (p ≥.05), was conducted to achieve the most parsimonious path analysis model. Unless previously specified otherwise, the criterion for statistical significance was p <.05 for each statistical test.

RESULTS

Sample characteristics

The demographic and clinical characteristics of the sample are presented in Table 1. Participants were mostly middle-aged, white spouses or adult children of a CCI patient. In this sample, a majority (71%) of participants did not consider themselves as an informal caregiver for the patient and 58% reported that they did not have prior experience with a loved one hospitalized in an intensive care unit. Although participants were expected to take part in medical decision making on behalf of a CCI patient, a moderate proportion (76%) of participants were responsible for making medical decisions for patients without an AD (LW or DPA) documentation in the patient’s medical records. There was no difference in decision burden (role stress and depression scores) when compared by AD status.

Table 1.

Demographics and Clinical Characteristics of Surrogate Decision Makers (n =489)

n Frequency (no.) Percentage (%)
Age (years, M ± SD) 483 53.1 14.3
Gender: female 489 370 76
Race: White 489 338 69
Marital status: Married 488 345 71
Caregiver status: Yes 486 141 29
Prior ICU exposure: Yes 486 198 42
Relationship to patient 489
 Spouse 180 37
 Adult child 138 28
 Other (DPA, other relative, non-relative) 171 35
Employment status: Employed 488 284 58
Annual household income 447
 ≤ $20, 999 103 23
 $21,000 to $49, 999 179 40
 ≥ $50, 000 165 37
Educational level 481
 < High school education 43 9
 High school education 57 68
 Undergraduate and/or graduate studies 20 23
Living will status: Yes 489 85 17
DPA status: Yes 489 98 20
Open in a new tab

Note. ICU = intensive care unit, DPA= durable power of attorney

Comparison of SDM characteristics by AD status

As shown in Table 2, there are several demographic and clinical characteristics of SDMs that differ by AD status. In particular, participants who were nonwhite, with an educational level of high school or less, and an annual household income less than $50,000 were more likely to make medical decisions for patients without ADs. The distribution of relatedness to patient was significantly different between SDMs with and without an AD. In particular, spouses and adult children were likely to have a CCI patient with an AD.

Table 2.

Comparison of Demographic and Clinical Characteristics of Surrogate Decision Makers by Advance Directive Status

Advance Directives Status
AD
No AD
Variable n M SD n M SD t
Age (in years) 115 55.2 13.8 368 52.4 14.5 −1.82
Role stress 41 55.4 32.8 141 60.9 30.5 .996
Depressive symptoms 109 23.3 11.9 357 23.6 11.2 .246
n % n % χ2
Race
 White 96 83 242 65 14.5**
 Non-white 19 17 132 35
Relationship to patient
 Spouse 44 38 136 36 13.0**
 Adult child 45 39 93 25
 Other (sibling, cousin, friend or DPA) 26 23 145 39
Annual household income
 ≤$20,999 17 16 86 25 7.89*
 $21,000 to $49,999 37 36 142 41
 ≥$50,000 50 48 115 34
Educational level
 < High school education 9 8 34 9 11.7**
 High school education 64 57 261 71
 Undergraduate or graduate education 40 35 73 20
Prior ICU experience:
 Yes 37 34 161 44 4.07*
 No 73 66 201 56
Caregiver status:
 Yes 46 40 95 74 8.83**
 No 69 60 276 26
Open in a new tab

Note. DPA: durable power of attorney; ICU: intensive care unit.

*

p<.05;

**

p<.01

Construction and validation of the path model

Construction of the path model

Based on the correlation coefficients presented in Table 3 and a review of the literature, the initial path model, shown in Figure 1a, was constructed and evaluated for validity in this sample of SDMs. Guided by O’Connor’s decision support framework (O’Connor et al., 1998) and a body of literature on advance care planning (Teno, et al., 2007; Vig, et al., 2007) as well as the psychological impact of critical illness on families (Douglas, et al., 2012; Douglas, Daly, O’Toole, & Hickman, 2009; Hickman, et al., 2010; Hickman, et al., 2012; Hickman & Douglas, 2010; Nelson, et al., 2007), the initial path model was constructed to underscore the predicative associations between SDM characteristics (age, gender, and race), the presence or absence of an AD, on the two measures of decisional burden (the SDM’s perception of role stress and depressive symptoms). As shown in Figure 2a, the initial path model illustrates direct and indirect relationships for SDM characteristics on role stress and depressive symptoms, and underscores that AD status has a direct relationship on the SDM’s perception of role stress and indirectly influences their appraisals of depressive symptoms.

Table 3.

Summary of Correlation Coefficients for Clinical Characteristics, Advance Directives, and Decisional Burden

Variable n 1 2 3 4 5 6 7 M SD
1. SDM Gender 489 — .03 .08* −.02 −.02 −.05 −.13*** — —
2. SDM Age 483 — .14*** .11** .08* −.08 −.11** — —
3. SDM Race 489 — .17*** .16*** −.04 .02 — —
4. Living will 489 — .67*** −.12* −.07 — —
5. DPA 489 — −.07 .01 — —
6. Role Stress 182 — .43*** 59.6 31.0
7. Depressive Symptoms 466 — 23.6 11.3
Open in a new tab

Note. SDM = surrogate decision maker; SDM Gender (0 = female, 1 = male); SDM Age = age in years; DPA = durable power of attorney (0 = no, 1 = yes); SDM race (0 = white, 1 = non-white); Living will (0 = no, 1 = yes). Two-tailed correlation coefficients are reported.

*

p<.10,

**

p<.05,

***

p<.01

Figure 1.

Figure 1

Open in a new tab

Participant flow chart. The database consisted of a 489 participants enrolled in the parent investigation. Of these 489 participants, 37% (n=182) of participants completed the decisional burden measures, which was used to examine the validity of hypothesized path model. SDMs = surrogate decision makers and CCI = chronically critically ill.

Figure 2.

Figure 2

Open in a new tab

Path analysis models. (A) Initial path analysis model with all hypothesized direct and indirect paths included, and (B) final path analysis model with all path coefficients statistically significant (p<.05). Standardized path coefficients are provided above each path and the explained variance of each endogenous variable is provided in bold. SDM: surrogate decision maker. DPA: durable power of attorney.

Validation of the path model

Once the initial path model was constructed, path analysis was conducted to assess the validity of the path model in this sample of SDMs. Although the initial hypothesized path analysis model had sufficient goodness-of-fit indices (χ2 = 13.3, df = 8, p = .10, CFI = .987, TLI = .954, RMSEA = .034), there were several path coefficients that did not meet the criterion for statistically significance. In order to identify the most parsimonious path analysis model, each insignificant path was independently removed. Based on a priori trimming procedures, paths that did not meet the retention criteria were individually removed from the analysis and the goodness-of-fit indices after removing each path are reported in Table 4. After the removal of 5 insignificant paths, the final path analysis model retained an excellent fit to these data (χ2 = 17.1, df = 13, p = .197, CFI = .990, TLI = .979, RMSEA = .023) and established the validity of the path analysis model.

Table 4.

Path Analysis: Path Removal and Goodness of Fit Indices

Path Analysis Model χ2 df p CFI TLI RMSEA
1. Initial 13.3 8 .100 .987 .954 .034
2. Removed path between SDM Race and Role stress 13.3 9 .147 .989 .967 .029
3. Removed path between DPA and Role stress 13.4 10 .201 .992 .977 .024
4. Removed path between SDM Gender and Role stress 13.9 11 .240 .993 .982 .021
5. Removed path between SDM Race and Depressive symptoms 15.7 12 .206 .991 .979 .023
6. Removed path between SDM Age and Depressive symptoms 17.1 13 .197 .990 .979 .023
Open in a new tab

Note. n =182, SDM: surrogate decision maker; DPA: durable power of attorney; CFI: comparative fit index; TLI: Tucker Lewis index; RMSEA: root mean square error of approximation

As shown in Figure 2b, the final path analysis model accounted for 4% of the explained variance in role stress and 20% of the explained variance in depressive symptoms. The SDM’s age (β = −.14, p = .02) and documentation of a living will in the patient’s medical records (β = −.13, p = .007) were two statistically significant direct predictors that reduced the participant’s appraisal of role stress. The second endogenous variable, depressive symptoms, was predicted by role stress (β = .43, p < .001) and the SDM’s gender (β = −.13, p = .04) while adjusting for the influence of documentation of a living will. In addition, there were four statistically significant correlational paths between demographic and clinical characteristics and AD (Figure 2b).

DISCUSSION

Chronic critical illness is anticipated to annually affect more than 600,000 Americans by year 2020 (Zilberberg, De Wit, et al., 2008; Zilberberg, de Wit, & Shorr, 2012; Zilberberg, Luippold, et al., 2008) and represents an emerging worldwide dilemma associated with growing populations of aging adults and advances in critical care therapeutics (Carson, Cox, Holmes, Howard, & Carey, 2006; MacIntyre, 2012; Nelson, et al., 2010). This life-limiting syndrome is associated with an increased incidence of cognitive impairment among affected patients and exposes family members and other designated persons to the burden of surrogate decision making. In the United States, federal legislation underscores the importance of establishing ADs to foster patient-centered decision making and alleviate the decisional burden associated with making complex healthcare decisions. However, most patients are affected by chronic critical illness will not have established an ADs or had a prior discussion with their designated surrogate decision maker about their preferences for life-sustaining treatments. This secondary data analysis aimed to describe the demographic characteristics of SDMs of patients with and without an AD and validate a path model that specifies the influence of ADs on the perception of decisional burden among SDMs of CCI patients.

First, there were several differences in demographic characteristics that varied among SDMs of patients with and without an established AD. In the U.S., the AD completion rates continue to increase; however, in this sample only 24% of CCI patients had documentation of an AD and confirm that majority of the SDMs were expected to make healthcare decision limited understanding of the patient’s wishes towards life-sustaining treatments. Notably, SDMs demographics, such as being white, a spouse or adult child with at least a high school education, were associated with the presence of an AD among CCI patients. Additionally, the presence of an AD was associated with SDMs who had a prior experience with a loved one in an intensive care unit and identified themselves as informal caregivers for a CCI patient. The extant literature on advance care planning interventions and informal caregiver research further support our findings regarding the clinical and demographic differences found in this sample of SDMs and CCI patients (Camhi, et al., 2009; Silveira, et al., 2010; Teno, et al., 2007; Vig, et al., 2007). Thus, the differences in clinical and demographic characteristics found in this study are well-established and consistent with the body of literature on AD completion rates across clinical populations of older adults.

Second, construction and validation of a path model was conducted to assess the predictive associations among SDM characteristics and ADs (living will and durable power of attorney) on measures of decisional burden. As shown in Figure 2b, the final path model established that the presence of a living will and the SDM’s age directly predicted the severity of the SDM’s role stress. Additionally, the SDM’s gender and role stress were direct predictors of the SDM’s depressive symptoms. These findings confirm that, when adjusting for the SDM’s gender, role stress mediates the relationship between AD status (presence of a living will) and the severity of depression among SDMs of CCI patients. The previous research (Silveira, et al., 2010; Teno, et al., 2007; Vig, et al., 2007) have examined the impact of ADs on family and patient outcomes and found that decisional burden was lessen when patients had an AD. However, this study is first to establish a mechanism through which ADs serve as an effective form of decision support that can mitigate the decisional burden associated with surrogate decision making for CCI patients. These results can used to develop efficacious decision support interventions aimed at attenuating the severity of role stress and depressive symptoms among SDMs of CCI patients.

Limitations

There are several limitations that affect the internal and external validity of this research. First, this study is a secondary data analysis with a cross-sectional design. Inherent to secondary data analysis, the investigators were limited by the variables contained within the dataset of the parent investigation, which constrained the dimensionality of decisional burden to measures of role stress and depression. Second, the cross-sectional design of this study does not permit evaluation of causal relationships across time. Lastly, the composition of the clinical and demographic characteristics of this convenience sample restricts the external generalizability of these findings to white, females who were the spouse or adult child of a CCI patient.

CONCLUSIONS

Chronic critical illness is recognized as emerging clinical population who require complex treatment decisions, which include life-sustaining treatments, palliative, and end-of life care. Advance directives are a primary prevention strategy that works to alleviate decisional burden. However, Americans are reluctant to preemptively complete these legal documents. As a result, a majority of family members are thrust into the SDM role and do not have knowledge of the patient’s preferences towards life-sustaining treatments, palliative, and end-of-life care. In particular, this study identified that SDMs who were non-white and had lower educational or socioeconomic status were associated with patients without an AD. Yet, the presence of an AD, particularly a living will, helped to attenuate the decisional burden associated with surrogate decision making for a CCI patient. Further exploration is needed to assess the longitudinal effects of ADs on the SDM’s decisional burden and evaluating opportunities to intervene with context specific and tailored decision support for SDMs of patients who do not have established ADs.

RELEVANCE TO CLINICAL PRACTICE

Most SDMs of chronically critically ill patients will not have the benefit of knowing the patient’s preferences for life-sustaining therapies and consequently be at high risk decisional burden. The results of this study are clinically useful and underscore the need for patient and family education on advance care planning and offer an opportunity to complete an AD before the patient is discharged from the hospital.

Relevance to clinical practice.

Study results are clinically useful for patient education on the influence of advance directives. Patients may be informed that SDMs without advance directives are at risk for increased decisional burden and will require decisional support to facilitate patient-centered decision making.

SUMMARY BOX.

What does this paper contribute to the wider global clinical community?

  • Establishing an advance directive or having discussions about life-sustaining treatments may reduce decisional burden and improve the quality of shared decision making.

  • Advance directives assist with the delivery of patient-centered care and reduce the psychological risks for surrogate decision makers.

Acknowledgments

Funding disclosure:

This publication was made possible by grants L30RR033212, L30MH091738, 2KL2TR000440, and R01NR008941 from the former National Center for Research Resources (NCRR), the National Center for Advancing Translational Science (NCATS), the National Institute of Mental Health (NIMH), and the National Institute of Nursing Research (NINR), components of the National Institutes of Health (NIH). Its contents are solely the responsibility of the authors and do not necessarily represent the official view of the NCRR, NCATS, NIMH, NINR, or the NIH.

Contributor Information

Ronald L. Hickman, Jr., Assistant Professor of Nursing, Case Western Reserve University Acute Care Nurse Practitioner, University Hospitals Case Medical Center, Department of Anesthesiology and Perioperative Medicine.

Melissa D. Pinto, Instructor of Nursing, Case Western Reserve University.

References

  1. Barry MJ, Edgman-Levitan S. Shared Decision Making — The Pinnacle of Patient-Centered Care. New England Journal of Medicine. 2012;366(9):780–781. doi: 10.1056/NEJMp1109283. [DOI] [PubMed] [Google Scholar]
  2. Bentler P. Comparative fit indices in structural models. Psychological Bulletin. 1990;107(2):238–246. doi: 10.1037/0033-2909.107.2.238. [DOI] [PubMed] [Google Scholar]
  3. Bentler P, Bonnett D. Significance tests and goodness of fit in the analysis of covariance structures. Psychological Bulletin. 1980;88(3):588–606. [Google Scholar]
  4. Brown M, Cudeck R. Alternatives ways of assessing model fit. In: Bollen KA, Long JS, editors. Reading and understanding multivariate statistics. Washington, DC: Americal Psychological Association; 1995. [Google Scholar]
  5. Camhi S, Mercado A, Morrison S, Du Q, Platt D, August G, Nelson J. Deciding in the dark: Advance directives and continuation of treatment in chronic critical illness. Critical Care Medicine. 2009;37(3):919–925. doi: 10.1097/CCM.0b013e31819613ce. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Carson SS, Cox CE, Holmes GM, Howard A, Carey TS. The changing epidemiology of mechanical ventilation: a population-based study. Journal of intensive care medicine. 2006;21(3):173–182. doi: 10.1177/0885066605282784. [DOI] [PubMed] [Google Scholar]
  7. Chelluri L, Im KA, Belle SH, Schulz R, Rotondi AJ, Donahoe MP, Pinsky MR. Long-term mortality and quality of life after prolonged mechanical ventilation. Critical Care Medicine. 2004;32(1):61–69. doi: 10.1097/01.CCM.0000098029.65347.F9. [DOI] [PubMed] [Google Scholar]
  8. Choi J, Sherwood PR, Schulz R, Ren D, Donahoe MP, Given B, Hoffman LA. Patterns of depressive symptoms in caregivers of mechanically ventilated critically ill adults from intensive care unit admission to 2 months postintensive care unit discharge: A pilot study*. Critical Care Medicine. 2012;40(5):1546–1553. doi: 10.1097/CCM.1540b1013e3182451c3182458. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Cox C, Carson S, Lindquist J. Differences in one year health outcomes and resource utilization by definition of prolonged mechanical ventilation: a prospective cohort study. Critical Care Medicine. 2007;11(1):120–131. doi: 10.1186/cc5667. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Cox C, Martinu T, Sathy S, Clay A, Chia J, Gray A, Tulsky J. Expectations and outcomes of prolonged mechanical ventilation. Critical Care Medicine. 2009;37(11):2888–2894. doi: 10.1097/CCM.0bo13e3181ab86ed. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Daly B, Douglas S, O’Toole E, Gordon N, Hejal R, Peerless J, Hickman R. Effectiveness trial of an intensive communication structure for families of long-stay ICU patients. Chest. 2010 doi: 10.1378/chest.10-0292. [Epub ahead of print] [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Douglas S, Daly B. Impact of a disease Management program upon caregivers of Chronically Critically Ill Patients. Chest. 2005;128(6):3925–3936. doi: 10.1378/chest.128.6.3925. [DOI] [PubMed] [Google Scholar]
  13. Douglas S, Daly B, Kelley C, O’Toole E, Montenegro H. Chronically Critically Ill Patients: Health-Related quality of Life and Resource Use after a Disease Management Intervention. Journal of Critical Care. 2007;16(5):447–457. [PMC free article] [PubMed] [Google Scholar]
  14. Douglas S, Daly B, Lipson A. Neglect of quality-of-life considerations in intensive care unit family meetings for long-stay intensive care unit patients*. Critical Care Medicine. 2012;40(2):461–467. doi: 10.1097/CCM.1090b1013e318232d318238c318234. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Douglas S, Daly B, O’Toole E, Hickman R. Depression among white and nonwhite caregivers of chronically critically ill. J Crit Care. 2009 doi: 10.1016/j.jcrc.2009.09.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Douglas S, Daly B, O’Toole E, Kelley C, Montenegro H. Age differences in survival outcomes and resource use for chronically critically ill patients. Journal of Critical Care. 2009;24(2):302–310. doi: 10.1016/j.jcrc.2008.02.004. http://dx.doi.org/10.1016/j.jcrc.2008.02.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Ehlenbach W, Hough C, Crane P, Haneuse S, Carson S, Curtis R, Larson E. Association between acute care and critical illness hospitalization and cognitive function in older adults. JAMA. 2010;303(8):763–770. doi: 10.1001/jama.2010.167. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Hickman R, Daly B, Douglas S, Clochesy J. Informational coping style and depressive symptoms in family decision makers of the cognitively impaired chronically critically ill. American Journal of Critical Care. 2010;19(5):410–420. doi: 10.4037/ajcc2010354. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Hickman R, Daly B, Lee E. Decisional conflict and regret: Consequences of surrogate decision making for the chronically critically ill. Applied Nursing Research. 2012;25(4):271–275. doi: 10.1016/j.apnr.2011.03.003. http://dx.doi.org/10.1016/j.apnr.2011.03.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Hickman R, Douglas S. Impact of chronic critical illness on the psychological outcomes of family members. AACN Advanced Critical Care. 2010;21(1):80–90. doi: 10.1097/NCI.0b013e3181c930a3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Joling KJ, Smit F, van Marwijk HWJ, van der Horst HE, Scheltens P, Schulz R, van Hout HPJ. Identifying target groups for the prevention of depression among caregivers of dementia patients. International Psychogeriatrics. 2012;24(02):298–306. doi: 10.1017/S1041610211001633. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Kline R. Principles and practices of structural equation modeling. 3rd. New York, NY: The Guildford Press; 2011. [Google Scholar]
  23. Kon A. THe shared decision-making continuum. JAMA. 2010;304(8):903–904. doi: 10.1001/jama.2010.1208. [DOI] [PubMed] [Google Scholar]
  24. Kramer BJ, Kavanaugh M, Trentham-Dietz A, Walsh M, Yonker JA. Predictors of Family Conflict at the End of Life: The Experience of Spouses and Adult Children of Persons with Lung Cancer. The Gerontologist. 2010;50(2):215–225. doi: 10.1093/geront/gnp121. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. MacIntyre NR. Chronic Critical Illness: The Growing Challenge to Health Care. Respiratory Care. 2012;57(6):1021–1027. doi: 10.4187/respcare.01768. [DOI] [PubMed] [Google Scholar]
  26. Marik P. Handbook of Evidence-Based Critical Care. Springer; New York: 2010. Chronic Critical Illness; pp. 43–54. [Google Scholar]
  27. Nelson J, Cox C, Hope A, Carson S. Chronic Critical Illness. American journal of respiratory and critical care medicine. 2010;182(4):446–454. doi: 10.1164/rccm.201002-0210CI. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Nelson J, Mercado A, Camhi S, Tandon N, Wallernstein S, August G, Morrison R. Communication about chronic critical illness. Archives of Internal Medicine. 2007;167(22):2509–2515. doi: 10.1001/archinte.167.22.2509. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. O’Connor A, Jacobsen M. Decisional conflict: supporting people experiencing uncertainty about options affecting their health. monograph. 2007 Retrieved from https://decisionaid.ohri.ca/ODST/pdfs/DC_Reading.pdf.
  30. O’Connor A, Jacobsen M, Stacey D. An evidence-based approach to managing women’s decisional conflict. JOGNN. 2005;31(5):570–581. doi: 10.1111/j.1552-6909.2002.tb00083.x. [DOI] [PubMed] [Google Scholar]
  31. O’Connor A, Legare F, Stacey D. Risk communication in practice: the contribution of decision aids. BMJ. 2005;327:36–40. doi: 10.1136/bmj.327.7417.736. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. O’Connor A, Tugwell P, Wells G, Elmslie T, Jolly E, Hollingworth G, Drake E. A decision aid for women considering hormone therapy after menopause: Decision support framework and evaluation. Patient education and counseling. 1998;33(3):267–280. doi: 10.1016/s0738-3991(98)00026-3. [DOI] [PubMed] [Google Scholar]
  33. Radloff LS. The CES-D Scale: A Self-Report Depression Scale for Research in the General Population. Applied Psychological Measurement. 1977;1(3):385–401. doi: 10.1177/014662167700100306. [DOI] [Google Scholar]
  34. Schulman RC, Mechanick JI. Metabolic and Nutrition Support in the Chronic Critical Illness Syndrome. Respiratory Care. 2012;57(6):958–978. doi: 10.4187/respcare.01620. [DOI] [PubMed] [Google Scholar]
  35. Silveira MJ, Kim SYH, Langa KM. Advance Directives and Outcomes of Surrogate Decision Making before Death. New England Journal of Medicine. 2010;362(13):1211–1218. doi: 10.1056/NEJMsa0907901. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Teno JM, Gruneir A, Schwartz Z, Nanda A, Wetle T. Association Between Advance Directives and Quality of End-of-Life Care: A National Study. Journal of the American Geriatrics Society. 2007;55(2):189–194. doi: 10.1111/j.1532-5415.2007.01045.x. [DOI] [PubMed] [Google Scholar]
  37. Tucker LR, Lewis C. A reliability coefficient for maximum likelihood factor analysis. Psychometrika. 1973;38:1–10. [Google Scholar]
  38. Vig E, Starks H, Taylor J, Hopley E, Fryer-Edwards K. Surviving Surrogate Decision-Making: What Helps and Hampers the Experience of Making Medical Decisions for Others. Journal of General Internal Medicine. 2007;22(9):1274–1279. doi: 10.1007/s11606-007-0252-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Zilberberg M, De Wit M, Pirone J, Shorr A. Growth in adult prolonged acute mechanical ventilation: Implication for healthcare delivery. Critical Care Medicine. 2008;36(5):1451–1455. doi: 10.1097/CCM.0b013e3181691a49. [DOI] [PubMed] [Google Scholar]
  40. Zilberberg M, de Wit M, Shorr A. Accuracy of previous estimates for adult prolonged acute mechanical ventilation volume in 2020: Update using 2000–2008. Critical Care Medicine. 2012;40(1) doi: 10.1097/CCM.0b013e31822e9ffd. [DOI] [PubMed] [Google Scholar]
  41. Zilberberg M, Luippold R, Sulsky S. Prolonged acute ventilation, hospital resource utilization, and mortality in the United States. Crit Care Med. 2008;36(3):724–730. doi: 10.1097/CCM.0B013E31816536F7. [DOI] [PubMed] [Google Scholar]

RESOURCES