Modifiable Factors Associated with the Completion of Advance Treatment Directives in Hematologic Malignancy: A Patient

Abstract

Objective:

The completion rates of advance treatment preferences in patients with hematologic malignancies are low. To improve these rates, the modifiable factors associated with completion need to be determined. This study aimed to examine the associations of patient attitudes toward, and knowledge about, advance directives (ADs) with the patient–caregiver dyadic completion of advance treatment directive surveys.

Methods:

Using a nonexperimental correlational design, 44 patient–caregiver dyads completed the questionnaires, including a Korean-Advance Directive model. Cohen's kappa coefficient and multiple logistic regression analyses examined the extent of dyadic agreement and patient factors for the dyadic completion of the advance treatment directive survey, respectively.

Results:

A minor group of patients (4.5%–11.4%) and caregivers (11.4%–18.2%) preferred aggressive end-of-life treatments, whereas more patients (47.7%) and caregivers (68.2%) supported hospice care. The only significant patient–caregiver dyadic concordance on treatment directives was for chemotherapy with a moderately high agreement (kappa = 0.60: 95% CI: 2.51–3.73). One score increase in AD knowledge and having a history of hematopoietic stem cell transplant (HSCT) increased the likelihood of dyadic completion of the treatment directive survey by 43% (p = 0.039) and 917% (p = 0.047), respectively.

Conclusions:

The patient–caregiver dyads in the setting of hematologic malignancy had a moderately high concordance with chemotherapy but were not associated with other treatment options. A higher level of AD knowledge and HSCT were associated with dyadic completion of the AD survey. Educational support is important to increase knowledge regarding ADs through ongoing palliative discussions among hematologic patients and their caregivers.

Introduction

The clinical benefits of the early introduction of advance care planning (ACP) and advance directive (AD) documentation in cancer have been well documented.^1,2 Thus, the integration of early palliative communication into standard cancer management is highly recommended to provide concomitant therapeutic and palliative care.^3–6 For example, depending on the patient circumstances, either a primary care physician or palliative care specialist begins a wide spectrum of palliative communication, including discussions about prognosis, the advantages and disadvantages of various therapeutic options, and further end-of-life (EOL) care issues, such as life-sustaining treatment (LST) preferences. Although the disease progresses, early and continuous integration of palliative communication may be more likely to lead to a smoother transition in care among the relevant specialist palliative care team members and put more weight on discussing EOL issues. Early adoption of palliative communication facilitates a seamless transition, which promotes AD documentation based on shared decisions, and ensures the continuity and high quality of EOL care that is in line with patient preferences.^3–6

However, access to such care is limited even in advanced cancer, with approximately less than half of patients with cancer having documentation of an AD.^7–9 Furthermore, given that patients and their relatives may have different preferences, an implementation of the documented AD reflecting the patient's preferences may sometimes be challenging, such that the individual's EOL needs cannot be met and high-quality care cannot be provided.^9,10 Absence of ACP and/or ADs or delayed implementation of these in cancer patients has led to poorer clinical outcomes, such as higher in-hospital deaths,^11–13 more resource utilization near death,^8,12,14 and delayed transition to palliative or hospice care.¹¹ The utilization of an AD is particularly challenging in patients with a hematologic malignancy,¹⁵ compared with those with metastatic solid tumors.¹⁶

One of the major barriers to such suboptimal use in patients with hematologic malignancies is poor prognostication.^11,17,18 Uncertainty in prognosis makes the decision making regarding the transition from the termination of curative therapeutics to palliative care challenging,^8,11,19 which then results in suboptimal utilization of AD and/or correspondingly leads to higher in-hospital deaths in this population.^8,9,11,19 Thus, EOL treatments often continue near death in patients with hematologic malignancies. For example, although aggressive advances in therapeutic treatments, such as hematopoietic stem cell transplantation (HSCT), successfully induce remission or prolong survival, these aggressive treatments often create challenges in making a timely decision for the termination of a curative intent of such care.¹¹ Timely decision making, which can be challenged by poor communication between the health care providers and the patients/family members about the prognostic uncertainty, leads to family members reporting that prognostic discussion facilitates their prognostic awareness of the patients' EOL, helps communication, and increases satisfaction with EOL care.²⁰

Given the suboptimal use of ADs, considerable adverse effects of no or suboptimal use of ADs, and the several challenges preventing the discussion and documentation of ADs, it is critical to find effective ways of optimizing the use of ADs. To effectively improve the utilization of ADs in patients with hematologic malignancies, both the patients and their families should be informed of ADs. Identification of influential factors associated with the utilization of ADs is also important. In the literature, some modifiable patient factors, such as attitudes toward or knowledge about ADs, are associated with enhancing individuals' autonomous EOL care decision making, and educational support for enhanced knowledge increases positive attitudes and more likely results in desiring an ACP and/or signing an AD.^12,21–23 In addition, reluctance to or fear of having such a discussion or communication difficulty is also associated with utilization of ADs, leading to lack of agreement to an ACP or AD between a patient and their caregiver.^24–26

To improve the documentation and utilization of ADs according to individual patients' preferences, agreement about treatment directives between patients and their caregivers is critical. Thus, factors affecting documentation and/or utilization of ADs between patient–caregiver dyads should be examined. However, those factors have been scantly addressed in cancer, particularly in hematologic malignancy, using both samples of patients and their caregivers. Therefore, this study examined the extent of agreement on advance treatment directives between the dyads of patients with hematologic malignancies who were admitted for HSCT and their caregivers, and determined the association of patient attitudes and knowledge with the dyadic completion with both patients and caregivers who completed the advance treatment directives survey.

Methods

Design and procedure

Using a nonexperimental descriptive correlational design, 44 patient–caregiver dyads participated and completed the survey questionnaires. Patients with hematologic malignancies who received care from a university-affiliated hospital in South Korea during an admission for HSCT from March to September 2017, and their caregivers, were invited to participate in this study. This participating hospital offers a comprehensive range of medical services for ∼40,000 oncological patients every year, and among these patients, hematology diseases comprise a large proportion (47.0%). The institutional review board of the university-affiliated hospital approved the study.

Two nurses experienced in oncology care enrolled the patient–caregiver dyads, then conducted face-to-face interviews with consented dyads. To minimize the nurses' influence on dyadic responses to the survey questionnaire, we developed the study protocol to specify a basic concept of ACP and ADs, and survey questionnaires and these administration strategies. Adhering to this protocol, the nurses explained and introduced the ACP and ADs as a vehicle for patient and family member planning of future EOL care. The nurses then assisted the patients and their caregivers in the completion of a Korean-Advance Directive (K-AD) model for treatment directives, and measures of attitudes toward and knowledge about ADs. If any dyad rejected filling out the K-AD model, nurses asked for and recorded the reasons.

Subjects

Inclusion criteria for the patients were (1) adults aged 21 years or older, (2) having a diagnosis of a hematologic malignancy, (3) being aware of their cancer diagnosis, (4) being capable of decision making for their own medical care, and (5) having a desire to discuss ADs. Caregivers were eligible if they (1) reported being caregivers who provide assistance with disease management to the patients as a spouse, child, sibling, relative, or friend; (2) were aged 21 years or older; and (3) agreed to participate in this study with the patients. Exclusion criteria for both patients and their family members were (1) if either a patient or his/her family caregiver expressed a reluctance to having such a discussion or (2) a patient qualified for or had a referral to hospice and palliative care.

Measures

Patients and their family caregivers responded to the K-AD model separately to avoid influencing each other's responses. To evaluate their attitudes toward and knowledge about ADs, the dyads also separately completed the questionnaires.

Advance directives

The K-AD model, which consists of three sections—a value statement, treatment directives, and proxy designation—was used to assess participants' preferences for EOL treatment directives. EOL treatment directives include aggressive LSTs, such as CPR, ventilation support, hemodialysis, chemotherapy, and hospice care. This model was developed to propose a culturally feasible form of ADs for Korean adults with or without a terminally ill condition.²⁷ The developers then revised this model for use in both cancer and noncancer contexts to reflect a recent act on LSTs that specify aforementioned EOL treatment directives.²⁸ The feasibility of this model was documented among cancer and elderly populations in both clinical and nonclinical settings.^29,30 The dyadic completion was defined as when both the patients and their caregivers had provided their responses to the treatment directives.

AD attitudes

The extent of an individual's positive or negative attitudes toward ADs was assessed using the 16-item Advance Directive Attitude Survey (ADAS).³¹ The scale was constructed using a 4-point Likert response with the options: 1 = strongly disagree through 4 = strongly agree. Possible scores ranged from 16 to 64, with higher scores indicating more positive attitudes toward ADs. The reliability of both the original version and Korean version was adequate in malignancy and nonmalignancy contexts.^31–33

Knowledge

The level of knowledge about ADs was assessed using the following portions of the ACP survey: living will (LW) (5 items), health care proxy (HCP) (6 items), and LSTs (12 items).³⁴ The scale was constructed using binary response options, with a score of 1 when mentioning a key concept for an item and a score of 0 for not mentioning the key concept. Possible scores when combining all subscales of LW, HCP, and LSTs range from 0 to 23, with higher scores indicating higher knowledge about ADs.³⁴ Reliability was desirable, with Cronbach's alphas of 0.93, 0.95, and 0.87 for each subscale, respectively.³⁴

Demographic and/or clinical characteristics

Demographic information for both the patients and their caregivers included age, gender, marital status, educational level, employment, patient and caregiver relationship, and comorbid conditions evaluated by the Charlson Comorbidity Index (CCI).³⁵ For the patients, clinical data, types of hematologic malignancies, duration of illnesses, and types of HSCT were obtained by medical record reviews.

Statistical analysis

Descriptive statistics, including frequency, percentage, and mean and standard deviation, were conducted for demographic characteristics of the patients and their family caregivers. The extent of patient–caregiver dyadic agreement on K-AD treatment directives was determined by a Cohen's kappa coefficient. t Tests were performed to compare differences in AD attitudes and knowledge between the dyads who completed or did not complete the AD treatment directives on the K-AD model. A multiple logistic regression analysis using the enter method was conducted to examine the associations of patients' attitudes toward and knowledge about ADs with the patient–caregiver dyadic completion of treatment directives. Statistical analyses were performed using the Statistical Package for the Social Sciences³⁶ with a level of statistical significance set at a p-value <0.05.

Results

Fifty pairs of patients with hematologic malignancies and their caregivers were initially offered participation in this study; 44 dyads (88%) agreed to participate. Either patients or caregivers of six dyads declined to participate in this study because of a lack of knowledge about ADs (n = 3), decisional difficulty in and fear of decision making about EOL care (n = 1), or both (n = 2).

Table 1 presents the demographics of the final sample composed of 44 patients (age, 48.61 ± 10.78 years; men, 52.3%; mean education, 12.07 ± 3.00 years) and their corresponding 44 caregivers (age, 47.50 ± 12.44 years; men, 34.1%; education, 13.09 ± 2.68 years) who completed the questionnaires. Most patients were married (81.8%) and the caregiver relationships were largely spousal (86.4%), adult child (11.4%), or friend (2.3%). The median time after cancer diagnosis was eight months (Q1–Q3 = 4.25–19.75). The major diagnoses were leukemia (40.9%) (acute myeloid leukemia, 34.1%; acute lymphoblastic leukemia, 4.5%; chronic myeloid leukemia, 2.3%), followed by lymphoma (27.3%), multiple myeloma (20.5%), and myelodysplastic syndrome (11.4%). Less than half of the patients had HSCT (40.9%), including autologous (22.7%), allogenic (13.6%), or umbilical cord blood stem cell transplantation (4.5%).

Table 1.

Demographic and Clinical Characteristics of the Dyads of Patients with Hematologic Malignancies–Caregivers (N = 44 Dyads)

Characteristics	Patients		Caregivers
Characteristics	n (%)	Mean ± SD	n (%)	Mean ± SD
Age (years)		48.61 ± 10.78		47.50 ± 12.44
Gender
Female	21 (47.7)		29 (65.9)
Male	23 (52.3)		15 (34.1)
Marital status
Married	36 (81.8)		37 (84.1)
Job (yes)	11 (25.0)		26 (59.1)
Education (years)		12.07 ± 3.00		13.09 ± 2.68
Caregiver relationships
Spouse			38 (86.4)
Child			5 (11.4)
Friends			1 (2.3)
Comorbidity	2.00 (2.00–4.00)
Duration of disease (months)		8.0 (4.25–19.75)
Type of cancer
^*Leukemia	18 (40.9)
Lymphoma	12 (27.3)
MM	9 (20.5)
MS	5 (11.4)
HSCT
Autologous	10 (22.7)
Allogenic	6 (13.6)
UCBSCT	2 (4.5)
None	26 (59.1)
^**AD need (NRS)		75 (50–100)		100 (50–100)

Normally distributed numerical variables were presented by mean ± standard deviation and non-normally distributed numerical variables were presented by median (Q1–Q3).

Includes acute lymphoblastic leukemia (4.5%), acute myeloid leukemia (34.1%), and chronic myeloid leukemia (2.3%).

Number of responses: patients, n = 29; caregivers, n = 30.

AD, advance directive; HSCT, hematopoietic stem cell transplantation; MM, multiple myeloma; MS, myelodysplastic syndrome; NRS, 0–100 numerical rating system; UCBSCT, umbilical cord blood stem cell transplantation.

K-AD treatment directives in malignancy cancer patient–caregiver dyads

Among the five treatment options of K-AD, a small group of patients and caregivers (4.5%–11.4% and 11.4%–18.2%, respectively) preferred aggressive EOL treatments (CPR [6.8% and 18.2%, respectively], ventilation support [4.5% and 13.6%, respectively], hemodialysis [9.1% and 11.4%, respectively], and chemotherapy [11.4% and 18.2%, respectively]) (Table 2). In contrast, more patients (47.7%) and caregivers (68.2%) supported hospice care. Most dyadic agreements for treatment wishes were not significant (CPR: kappa = 0.06, 95% CI: 0.13–20.94; ventilation support: kappa = 0.07, 95% CI: 0.15–20.23; hemodialysis: kappa = 0.04, 95% CI: 0.12–4.21; hospice care: kappa = 0.04, 95% CI: 0.18–8.87). The only significant patient–caregiver dyadic consensus on treatment directives was for chemotherapy (kappa = 0.60, 95% CI: 2.51–3.73) indicating a moderately high level of concordance for chemotherapy among the dyads (Table 2).

Table 2.

The Degree of Dyadic Agreement with Treatment Wishes between Patients with Hematologic Malignancies and Their Caregivers

Treatment directive options	Patients (n = 34)	Caregivers (n = 36)	Kappa (p)	95% CI
Treatment directive options	Frequency (%)	Frequency (%)	Kappa (p)	Lower	Upper
CPR	3 (6.8)	8 (18.2)	0.06 (0.700)	0.13	20.94
Ventilator	2 (4.5)	6 (13.6)	0.07 (0.658)	0.15	20.23
HD	4 (9.1)	5 (11.4)	0.04 (0.829)	0.12	14.21
Chemotherapy	5 (11.4)	8 (18.2)	0.60 (0.001)	2.51	3.73
Hospice	21 (47.7)	30 (68.2)	0.04 (0.819)	0.18	8.87

Of 44 patient–caregiver dyads, 34 patients and 36 caregivers completed their desires for end-of-life treatment directives using the Korean-Advance Directive model; the remaining dyads declined to complete the model.

Bold indicates that a significant number of patients (47.7%) and more than half the caregives (68.2%) preferred hospice treatment.

CPR, cardiopulmonary resuscitation; HD, hemodialysis.

Factors associated with the patient–caregiver dyadic completion of treatment directives

The level of AD attitudes among patients with hematologic malignancy was moderately high with a mean score of 49.45 (±5.29); knowledge about ADs was substantially poor, with a median score of 3.00 (Q1–Q3 = 1.00–8.00). AD knowledge was significantly different between patients and their caregivers who both completed (completers) and those who did not complete (noncompleters) the EOL treatment directives, with completers having greater AD knowledge (1.23 vs. 3.97, t = 2.80, p = 0.008).

In a multiple logistic regression analysis, AD attitudes and AD knowledge were entered as independent variables for completion of the patient–caregiver dyadic EOL treatment directive; age, education level, comorbidities, cancer duration, and ≥1 HSCT experience were entered as covariates (Table 3). Knowledge and ≥1 HSCT experience increased the likelihood of the dyadic completion of EOL treatment directives. A one-score increase in patient knowledge was associated with a 43% increase in completing the EOL treatment directives (AOR = 1.43, p = 0.039; CI = 1.02–2.00). There was a 10-fold increase in dyadic completion of EOL treatment directives in patients with HSCT compared with those without HSCT (AOR = 10.17, p = 0.047; CI = 1.03–100.08). The overall goodness of fit was adequate (χ² = 14.23, p = 0.047; Nagelkerke R² = 0.39).

Table 3.

Factors Associated with the Patients with Hematologic Malignancies–Caregivers' Dyadic Completion of Treatment Directives (N = 44)

Patient factors	B	AOR	95% Confidence interval		p
Patient factors	B	AOR	Lower	Upper	p
Age	−0.04	0.97	0.89	1.05	0.385
Education (years)	0.07	1.07	0.81	1.43	0.631
Cancer duration (months)	−0.01	0.99	0.97	1.01	0.442
HSCT (yes)	2.32	10.17	1.03	100.08	0.047
Comorbidity	−0.09	0.91	0.65	1.30	0.611
AD attitudes	0.07	1.07	0.92	1.25	0.385
AD knowledge	0.36	1.43	1.02	2.00	0.039
Model summary	χ² = 14.23, p = 0.047, Nagelkerke R² = 0.39

Bold indicates that in a multiple logistic regression analysis, AD attitudes and AD knowledge were entered as independent variables for completion of the patient–caregiver dyadic EOL treatment directive; age, education level, comorbidity, and disease duration were entered as continuous independent variables, and HSCT was applied as dummy variables.

AD, advance directive; AOR, adjusted odds ratio; HSCT, hematopoietic stem cell transplantation.

Discussion

In this study, we explored agreement in the EOL treatment wishes in patients with hematologic malignancies using a sample of patient–caregiver dyads, a population that has been rarely investigated. The patient–caregiver dyads did not show significant agreement in most treatment wishes, except for chemotherapy, in which dyadic agreement was moderately high, which demonstrates the need for communication between patients and their caregivers regarding future treatment and care. Furthermore, knowledge about ADs was the modifiable factor significantly associated with the completion of AD survey. Thus, knowledge can be an important target by interventions that improve utilization of ADs in this population.

ACP and/or AD utilization most frequently targets patients with solid mass cancers, whereas less frequently targeting those with hematologic malignancies. Suboptimal use of such care in patients with hematologic malignancies was largely attributed to difficulties in decision making regarding the timely termination of therapeutic intent and transitioning to palliative care because of uncertainty in the disease prognosis.^8,11,17 Under the circumstances that utilization of ADs is suboptimal, particularly in hematologic malignancies,^8,11,17,18 and given the fact that little is known about knowledge and attitudes regarding ADs in Asian contexts,¹² overall responses of patients with hematologic malignancies and their caregivers to the AD survey indicated that both groups seemed to be agreeable to addressing ADs.

Either patients or caregivers of 6 (12%) out of 50 dyads declined study participation. Thirty-one out of the 44 dyads who participated in this study (70.4%) completed the K-AD treatment directives, which differs from the results of a prior study of solid mass cancers. In a previous study,³⁰ approximately half of the patients with solid cancers and their families declined to participate (44.4%) because of decisional difficulty regarding EOL care (50.0%), an apparent frustration and/or hopelessness after AD documentation (22.2%), and lack of knowledge about ADs (8.3%). In China, where ADs are not widely accepted among cancer patients and their family members, 40% (166/412 patients or their family members) signed an AD. However, only 24 patients actually participated in the AD decision making.³⁷ Thus, despite the uncertain prognosis of hematologic malignancies, Korean patients with hematologic malignancies and their families showed positive attitudes toward addressing ADs, which demonstrates that early introduction of ADs in this population is feasible.

Positive attitudes toward ACP were reported from both survivors of hematopoietic cell transplant and their caregivers who felt that ACP discussions were likely to reduce the burden (67%) without depriving them of hope.³⁸ The families of patients who died in hospice care or did not die in a hospital frequently reported better EOL care than those who did not receive hospice care, or received it for only a short time period, or who died in a hospital. These findings indicate support for ACP consistent with the preferences of the patients, including earlier hospice referral and avoidance of intensive care unit admissions and hospital death.¹³ Although there has been controversy and debate among stakeholders regarding the right time,³⁹ early adoption of ACP discussion and/or ADs across all stages of cancer is likely to promote patient and family understanding and knowledge about ACP and ADs. Engaging in AD discussion will eventually facilitate informed decision making for EOL care that will result in the delivery of EOL care according to the patient's wishes, ensuring patient and family satisfaction and the quality of EOL care.⁶

In our investigation of dyadic agreement with EOL preferences in dyads who completed this AD survey, a small group of both patients and caregivers had preferences for aggressive EOL treatments such as CPR, ventilator support, hemodialysis, or chemotherapy. However, our study indicated that the extent of patient–caregiver dyadic consensus on EOL treatment directives was poor, with significant moderately high concordance only for chemotherapy. Only 11.4% of patients preferred chemotherapy, and <10% preferred other aggressive treatments, whereas more caregivers (11.4%–18.2%) preferred aggressive treatments. On the contrary, approximately half of the patients (47.7%) and more than half of the caregivers (68.2%) in this study supported hospice care, but there was no significant concordance between the preferences. In a prior study of 44 patients with cancer and their families who completed the K-AD, their preferences for aggressive treatment showed mild to moderate concordance for CPR (kappa = 0.43, p = 0.004) and ventilator support (kappa = 0.28, p = 0.046).³⁰ Different opinions of patients and caregivers may lead to the poor concordance. For example, in patients with cancer or community-dwelling elders, avoiding family burden was one of the most important and common wishes.^29,30 Thus, patients might be less inclined to receive care, including aggressive treatment and hospice care, than their caregivers may wish for them. In the circumstances of poor patient–caregiver agreement on EOL care, a patient's decisional incapacity may lead the patient to receiving unwanted treatments until near death, against their own preferences but in accordance with their caregiver's wish. Unless preferred EOL care is communicated and documented among patients, their caregivers, and health care providers, EOL care decision making in South Korea can be taken by the family members near a patient's time of death,^40,41 even in the absence of the patient's input.⁴²

Thus, communication about ACP and/or ADs is important to provide the EOL care that the patients valued and desired or, at least, to reduce decisional conflicts among the stakeholders about an individual's EOL care, including LSTs. Although there has been controversy and debate among stakeholders about the right time,³⁹ early adoption of ACP discussion and/or ADs across all stages of cancer is likely to promote patient and family understanding and knowledge about ACP and ADs. In a qualitative hematologic malignancy study, most patients refused LSTs, whereas patients used ADs to inform their EOL wishes to their family, which implies that patients valued communication about their wishes with their family members.⁴³ According to a recent systematic review with meta-analysis that reported the impact of communication between health care professionals and nursing home residents or families on patient outcomes, EOL communication helped with a limit or withdrawal of LSTs.⁴⁴ Together, findings of this and prior studies supported the need for effective and timely communication about future EOL care between patients with cancer and their family members to reduce any decisional conflicts and burdens of families and provide adequate EOL care to patients.

Engaging in AD discussion will also facilitate informed and shared decision making for EOL care that will result in the delivery of EOL care according to the patient's wishes and promote the quality of EOL care.^6,44 Particularly in hematology malignancies characterized by difficult prognostication, smooth transition from curative- to palliative-oriented care is likely through such palliative discussion in which health care professionals discuss prognosis, therapeutic risks/benefits, and other therapeutic options, and assist patients and families with informed and shared decision making for EOL care. Otherwise, aggressive therapeutic care continued to near death is largely associated with prognosis uncertainty in hematology malignancy.^8,11 Continuation of aggressive treatment implies a curative intention while near death, which commonly results in palliative consultation delays; thus, only a small group of patients (16%–21%) receive consultation more than a month before death.^8,11 Therefore, health care professional training that includes communication skills empowers them to offer periodic EOL conversations during routine care.

Despite the association of ACP discussion and ADs to improvements in patient and clinical outcomes,^8,11–14,44 their access or implementation as documented in clinical practice remained inadequate,⁴⁵ particularly in hematologic malignancies in the outpatient setting.^17,44 Limited access to ACP and AD discussions often led patients with advanced hematologic malignancies to receive aggressive treatments, such as an intensive care unit or chemotherapy near death.^8,11 According to a recent report for the past decade in South Korea,⁴⁶ aggressive treatments that were delivered at the EOL remained high among terminal cancer patients with solid cancers, which shortened the time from the last chemotherapy to more resource utilization at the last month of life, such as intensive care unit care and emergency visits.

There were known barriers to the optimal use of such care, including prognostic uncertainty, communication difficulty, lack of professional training, time conflicts, and lack of a system or policy across health care agencies.^{8,10–11,14,41,47} However, some modifiable explanatory factors, such as attitudes toward or knowledge about ADs, could be the target of interventions, but have been little addressed. This study examined attitudes and knowledge about ADs and found that increased knowledge about ADs were associated with a greater likelihood of patient–caregiver dyadic completion of treatment directives. Dyad knowledge about ADs was substantially poor, although they showed moderately positive attitudes toward ADs. Given that survivors of HSCT and their caregivers believed that the ACP discussions were likely to reduce burden (67%) rather than deprive them of hope,³⁸ a positive effect of AD knowledge was a greater likelihood of dyadic completion of treatment directives, implying that knowledge can be an important contributor to improvement in the utilization of ADs in this population; however this implication requires validation in a larger sample.

Although most of the nonmodifiable factors, including age, which in past studies often affected the positive impact on the AD documentation or related care,¹⁷ were not significant, a disease-specific factor, a history of HSCT, was associated with an increased likelihood of dyadic completion of ADs by ∼10-fold in this study. Previously, demographic and clinical influences on ACP or AD documentation varied by health contexts, and collectively, older age, female gender, higher education, or family/caregiving factors were associated with desiring or having an AD or palliative care preference.^{12,17,48–50} In contrast with this study, an older age of transplant recipients (mean: 55 years) significantly increased the odds of AD completion.¹⁷ Among Chinese cancer patients whose AD knowledge was substantially poor, older age, female gender, a higher educational level, religious beliefs, and poorer functional status were associated with greater likelihood of desiring ADs.¹² Adults in Hong Kong ≥50 years claimed to demonstrate more preference for palliative care, whereas those who did not care for chronically ill family members had a lower preference.⁴⁹ Regarding the family/caregiving support factor, and a higher education (>12 years),⁴⁸ married/and widowed patients at the outpatient clinics were more likely to have an AD than single or separated patients. In this study, age was not significantly associated with AD documentation. A possible reason may be the fact that the mean ages of both patients and caregivers were relatively lower, which might prevent them from contemplating the need for ADs.

The association of HSCT with completion of ADs was inconsistent based on the limited available evidence in the literature.^8,17 In one study, allogenic HSCT recipients and their providers were less likely to utilize palliative care services or to engage in early ACP communication,⁸ whereas in another study of HSCT recipients, allogeneic transplant recipients were more likely to complete ADs than those with autologous HSCTs.¹⁷ In this study, the group of patients comprised non-HSCT recipients (59.1%) and HSCT recipients (40.9%; 18% of which were allogenic HSCT recipients), and any type of HSCT experience substantially increased the likelihood of dyadic completion of ADs. Thus, it remains unconfirmed whether the types of HSCT affect ACP and/or AD documentation, and further studies are warranted to examine this relationship more thoroughly.

Limitations

This study had several limitations. A single hospital-based small sample could impact the study results, requiring a validation of the results in a larger sample from multiple centers. Patients with different types of hematologic malignancies experience different progressions and will naturally have differing characteristics, such as age, which can affect patients' and their caregivers' willingness to complete an AD and their preferences for EOL treatments. However, this small sample precludes the comparisons of AD preferences and their AD-related behaviors by the types of hematologic malignancies. This study was also composed of relatively young patients and caregivers, which might result in nonsignificant associations of age with the dyadic completion of treatment directives. Thus, further studies using larger more diversely aged samples are needed to examine the relationship more thoroughly. Furthermore, our participants were inpatients who might have been more responsive to the topic; thus, their responses may not represent attitudes toward ADs by all patients with hematologic malignancies. Outpatients may provide a different insight into ADs and EOL treatments.

Conclusion

Utilizing ADs appears feasible in patients with hematologic malignancies, considering that approximately half our patient–caregiver dyads completed the AD survey. Patients with hematologic malignancies and their caregivers did not favor overaggressive treatments and did not show agreement in treatment options except for chemotherapy. Knowledge of ADs and HSCT experiences were associated with a greater likelihood of dyadic completion of ADs. Thus, educational support is important for increasing knowledge of ADs and decreasing decisional conflicts through ongoing palliative care discussions among patients with hematologic malignancies and their caregivers. This finding is preliminary but conveys the message that efforts are needed to narrow any discrepancies between the patients and their caregivers. To narrow the discrepancies and to improve AD utilization, patients and their families may need more information about ADs. Health care providers are responsible for assisting patients and their families by providing this information and reducing the EOL decisional conflicts to promote AD utilization based on informed decision making and shared goals of care.

This original study provides important insights and initiatives for the ACP and ADs embedded within standardized care in hematologic malignancies. This study also suggests how support for patients, and their caregivers' engagement in discussions and planning for EOL care, decrease EOL decisional conflicts and promote informed decision making and shared goals of care. Future research should prioritize studies on designing an educational intervention to increase patient and family knowledge that facilitates the translation into actual AD documentation. Additional studies are also warranted to design and evaluate palliative interventions with AD utilization based on the known modifiable and nonmodifiable factors.

Footnotes

Funding Information

No funding was received for this article.

Author Disclosure Statement

No competing financial interests exist.

References

Detering

, Silveira

. Advance care planning and advance directives. www.uptodate.com/contents/advance-care-planning-and-advance-directives (Last accessed May 04, 2018).

Narang

, Wright

, Nicholas

. Trends in advance care planning in patients with cancer: Results from a national longitudinal survey. JAMA Oncol, 2015; 1:601–608.

Epstein

, Morrison

. Palliative oncology: Identity, progress, and the path ahead. Ann Oncol, 2012; 23:43–48.

Ferrell

, Temel

, Temin

, et al. Integration of palliative care into standard oncology care: American Society of Clinical Oncology clinical practice guideline update. J Clin Oncol, 2017; 35:96–112.

Jang

, Sundaram

, Finlay

, et al. Early discussion of advance directives for patients with newly diagnosed cancer. J Clin Oncol, 2016; 34:91.

Ranganathan

, Gunnarsson

, Casarett

. Palliative care and advance care planning for patients with advanced malignancies. Ann Palliat Med, 2014; 3:144–149.

Brown

, Shen

, Urbauer

, et al. Room for improvement: An examination of advance care planning documentation among gynecologic oncology patients. Gynecol Oncol, 2016; 142:525–530.

Schoenbeck

, McNey

, Eckhert

, et al. Advance care planning and specialty palliative care utilization for patients with hematologic malignancies who undergo allogeneic hematopoietic cell transplant. J Clin Oncol, 2016; 34:18.

Schubart

, Levi

, Bain

, et al. Advance care planning among patients with advanced cancer. J Oncol Pract, 2019; 15:e65–e73.

10.

Carr

, Luth

. Advance care planning: Contemporary issues and future directions. Innov Aging, 2017; 28:1.

11.

Eckhert

, Schoenbeck

, Galligan

, et al. Advance care planning and end-of-life care for patients with hematologic malignancies who die after hematopoietic cell transplant. Bone Marrow Transplant, 2017; 52:929–931.

12.

Zheng

, Fu

, Xiang

, et al. Knowledge, attitudes, and influencing factors of cancer patients toward approving advance directives in China. Support Care Cancer, 2016; 24:4097–4103.

13.

Wright

, Keating

, Ayanian

, et al. Family perspectives on aggressive cancer care near the end of life. JAMA, 2016; 315:284–292.

14.

Rocque

, Dionne-Odom

, Sylvia Huang

, et al. Implementation and impact of patient lay navigator-led advance care planning conversations. J Pain Symptom Manage, 2017; 53:682–692.

15.

Freeman

, Wood

, Fox

, Hanson

. Access to palliative care consultation and advance care planning for adults with high-risk leukemia. J Palliat Med, 2018; 21:225–228.

16.

Hui

, Didwaniya

, Vidal

, et al. Quality of end-of-life care in patients with hematologic malignancies: A retrospective cohort study. Cancer, 2014; 120:1572–1578.

17.

Wang

, Ma

, Nelson

, et al. Advance care planning and palliative care integration for patients undergoing hematopoietic stem-cell transplantation. J Oncol Pract, 2017; 13:e721–e728.

18.

Wang

, Ma

, Nelson

, et al. Advance care planning and palliative care consultation for stem cell transplant patients. J Clin Oncol, 2016; 34:113.

19.

McCaughan

, Roman

, Smith

, et al. Determinants of hospital death in haematological cancers: Findings from a qualitative study. BMJ Support Palliat Care, 2018; 8:78–86.

20.

Krawczyk

, Gallagher

. Communicating prognostic uncertainty in potential end-of-life contexts: Experiences of family members. BMC Palliat Care, 2016; 15:59.

21.

Hamayoshi

. Effects of an education program to promote advance directive completion in local residents. General Med, 2014; 15:91–99.

22.

Hamayoshi

, Goto

, Matsuoka

, et al. Effects of an advance care planning educational programme intervention on the end-of-life care attitudes of multidisciplinary practitioners at an acute hospital: A pre- and post-study. Palliat Med, 2019; 33:1158–1165.

23.

Cattagni Kleiner

, Santos-Eggimann

, Fustinoni

, et al. Advance care planning dispositions: The relationship between knowledge and perception. BMC Geriatr, 2019; 19:118.

24.

Luck

, Eggenberger

, Newman

, et al. Advance directives in hospice healthcare providers: A clinical challenge. Am J Med, 2017; 130:e487–e489.

25.

Lund

, Richardson

, May

. Barriers to advance care planning at the end of life: An explanatory systematic review of implementation studies. PLoS One, 2015; 10:e0116629.

26.

Paladino

, Bernacki

, Hutchings

, et al. Facilitators and barriers to advance care planning in oncology. J Clin Oncol, 2014; 32:84.

27.

Kim

, Hong

, Kim

, et al. Feasibility evaluation of Korean advance directives (K AD). J Korean Acad Soc Nurs Educ, 2014; 20:639–649.

28.

National Law Information Center. Korean act (Act No. 14013) on the determination of life-sustaining treatment for terminally ill patients and palliative care. http://law.go.kr/lsEfInfoP.do?lsiSeq=180823 (Last accessed February 29, 2017)

29.

Kim

, Hong

, Kim

. Feasibility of the Korean-advance directives among community-dwelling elderly persons. Holist Nurs Pract, 2017; 31:234–242.

30.

Kim

, Koh

, Park

, Kim

. End-of-life care decisions using a Korean advance directive among cancer patient-caregiver dyads. Palliat Support Care, 2017; 15:77–87.

31.

Nolan

, Bruder

. Patients' attitudes toward advance directives and end-of-life treatment decisions. Nurs Outlook, 1997; 45:204–208.

32.

Kim

, An

, Heo

, Shin

. Attitudes toward advance directives and prognosis in patients with heart failure: A pilot study. Korean J Intern Med, 2019 [Epub ahead of print]; DOI: 10.3904/kjim.2018.158.

33.

Lee

, Park

. Factors influencing attitude toward advance directives of older cancer patients. Korean J Adult Nurs, 2015; 27:449–458.

34.

, Lee

. End-of-life communication: Ethnic differences between Korean American and non-Hispanic White older adults. J Aging Health, 2009; 21:967–984.

35.

Charlson

, Pompei

, Ales

, MacKenzie

. A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation. J Chronic Dis, 1987; 40:373–383.

36.

IBM Corp. IBM SPSS Statistics for Windows, Version 25.0. Seoul, South Korea: SPSS Korea, Inc., 2019.

37.

Xing

, Lin

.-X, Li

, et al. Advance directives: Cancer patients' preferences and family-based decision making. Oncotarget, 2017; 8:45391–45398.

38.

Loggers

, Lee

, Chilson

, et al. Advance care planning among hematopoietic cell transplant patients and bereaved caregivers. Bone Marrow Transplant, 2014; 49:1317–1322.

39.

Keam

, Yun

, Heo

, et al. The attitudes of Korean cancer patients, family caregivers, oncologists, and members of the general public toward advance directives. Support Care Cancer, 2013; 21:1437–1444.

40.

Kwak

, Salmon

, Acquaviva

, et al. Benefits of training family caregivers on experiences of closure during end-of-life care. J Pain Symptom Manage, 2007; 33:434–445.

41.

Kim

, Kim

, Lee

, KIM

. Perception towards advance directives (AD), hospice palliative care, life-sustaining treatment (LST) among middle aged adults. Korean J Soc Welf Res, 2013; 37:21–44.

42.

Heo

. Patient autonomy and advance directives in Korea. J Korean Med Assoc, 2009; 52:865–870.

43.

Trarieux-Signol

, Bordessoule

, Ceccaldi

, et al. Advance directives from haematology departments: The patient's freedom of choice and communication with families. A qualitative analysis of 35 written documents. BMC Palliat Care, 2018; 17:10.

44.

Gonella

, Basso

, Dimonte

, et al. Association between end-of-life conversations in nursing homes and end-of-life care outcomes: A systematic review and meta-analysis. J Am Med Dir Assoc, 2019; 20:249–261.

45.

Barakat

, Barnes

, Casanova

, et al. Advance care planning knowledge and documentation in a hospitalized cancer population. Proc (Bayl Univ Med Cent), 2013; 26:368–372.

46.

Choi

, Keam

, Kim

, et al. Cancer treatment near the end-of-life becomes more aggressive: Changes in trend during 10 years at a single institute. Cancer Res Treat, 2015; 47:555–563.

47.

Howell

, Wang

, Roman

, et al. Preferred and actual place of death in haematological malignancy. BMJ Support Palliat Care, 2017; 7:150–157.

48.

Alshanberi

ATC

, Huddleston

, Imam

, et al. Advance directives in patients over 60 years old: Assessment of perceived value and need for education in the outpatient setting. Arch Med, 2018 [Epub ahead of print]; DOI: 10.21767/1989–5216.1000277

49.

Chung

, Wong

, Kiang

, et al. Knowledge, attitudes, and preferences of advance decisions, end-of-life care, and place of care and death in Hong Kong. A population-based telephone survey of 1067 Adults. J Am Med Dir Assoc, 2017; 18:367 e19–367 e27.

50.

Hong

, Kwon

, Kim

, et al. Adopting advance directives reinforces patient participation in end-of-life care discussion. Cancer Res Trea, 2016; 48:753–758.

Modifiable Factors Associated with the Completion of Advance Treatment Directives in Hematologic Malignancy: A Patient–Caregiver Dyadic Analysis

Abstract

Objective:

Methods:

Results:

Conclusions:

Introduction

Methods

Design and procedure

Subjects

Measures

Advance directives

AD attitudes

Knowledge

Demographic and/or clinical characteristics

Statistical analysis

Results

K-AD treatment directives in malignancy cancer patient–caregiver dyads

Factors associated with the patient–caregiver dyadic completion of treatment directives

Discussion

Limitations

Conclusion

Footnotes

Funding Information

Author Disclosure Statement

References