Caregiver perceptions of end-of-life care in patients with high-grade glioma

Abstract

Background

Patients dying from high-grade gliomas (HGG) suffer from high symptom burden in the end-of-life (EoL) phase. Family caregivers are most informed about the patient’s symptoms and disease course. The aim of this study is to assess caregiver perception on quality of EoL care of HGG patients.

Methods

Caregivers prospectively participated in the Toolkit After-Death Bereaved Family Member Interview, part of the Toolkit of Instruments to Measure End-of-Life Care (TIME survey). This validated survey assesses EoL care in areas such as physical comfort and emotional support, advance care planning, focus on the individual, attention to family, and coordination of care. The quality of EoL care was measured by domain scores (0 = care was always optimal, 1 = care was always suboptimal) or with a 0-10 scale.

Results

Of the 55 enrolled family caregivers, 44 completed the interview and rated the overall care high (8.90 ± 1.36/10), perceived that patients’ wishes were respected (9.46 ± 0.95) and that they died in dignity (9.65 ± 0.98). Caregivers perceived high satisfaction with information and decision-making (0.18), advance care planning (0.19), focus on the individual (0.16), and care coordination (0.11). Attention to family (0.25) needed improvement. Only 41% of caregivers were confident that they knew what to do at the time of death and 46% felt that the healthcare team did not provide them with someone to turn to in distress.

Conclusions

Caregivers reported high overall satisfaction with EoL HGG care, though attention to family and communication needed improvement. Focus should therefore be on improved caregiver communication to improve EoL care, caregiver burnout, and bereavement in HGG populations.

High-grade gliomas (HGG) include anaplastic astrocytomas, anaplastic oligodendroglioma (World Health Organization [WHO] grade III), and glioblastomas (WHO grade IV). Treatment of HGG typically consists of surgical resection followed by radiation therapy with concurrent and adjuvant temozolomide.¹ Despite aggressive treatment, the prognosis is poor, with a median survival of 2-5 years and less than 15 months for WHO grade III and grade IV tumors, respectively.^2–4

Several studies have shown that patients dying from HGG suffer from a high symptom burden.^5,6 Systematic literature reviews show that there are very limited data describing the symptoms that HGG patients experience at the end of life (EoL). Prospective EoL studies are difficult to perform and therefore show wide ranges of symptoms. The most common symptoms include poor communication (64%-90%), drowsiness (1%-90%), seizures (3%-56%), dysphagia (7%-85%), and headaches (4%-62%).^5,6 These symptoms, along with impaired neurological integrity, can cause changes in mood, personality, and cognitive abilities and may raise unique needs at the EoL when compared to patients dying from non-neurological processes.^7–10 It is well documented that HGG patients report high levels of psychological needs and support, particularly when distressed.¹¹ Therefore, brain tumor patients might benefit from specific guidelines for EoL care,^6,12 and early advance care planning has been highly encouraged to anticipate for impaired decision-making skills in the later disease course.^5,12–14 A recent database analysis of 12 437 patients who died from malignant glioma revealed that only 63% of patients were enrolled in hospice. Of those enrolled, the median hospice course was 21 days.¹⁵ Hospice care and palliative care utilize a multidisciplinary approach to focus on quality of life issues and symptom management of patients with life-threatening diseases. While palliative care is typically delivered concurrently to active cancer treatment, hospice focuses on patients with a prognosis of 6 months or less. Patients enrolled into hospice programs are typically no longer seeking any curative or life-prolonging treatment but want to focus on symptom control and quality of life instead. Palliative care as well as hospice is based on a multidisciplinary team approach that includes nurses, a hospice physician, social workers, chaplains, bereavement counselors, and volunteers. Hospice care can be delivered wherever the patient lives (their home or that of someone close to them, assisted living facility, nursing home) or an inpatient facility if warranted for symptom management.^16,17 Hospice care has been shown to result in lower rates of hospitalizations, emergency department visits, and intensive care unit admissions of cancer patients in their last year of life.^18,19

Since HGGs in the EoL phase are predominantly managed in the outpatient setting, daily care of patients with HGG falls on the shoulders of family and informal caregivers. Caregivers and HGG patients have reported similarly poor quality of life²⁰ and high distress levels.²¹ Given the caregiver’s proximity to the HGG patient, their perception of patient care can offer unique insights. It is the aim of this study to assess the quality of EoL care provided to HGG patients through the perspective of the caregiver and to understand how caregivers might be better supported throughout the dying process.

Methods

This study was approved by the local Institutional Review Board. All patients and caregivers provided separate informed consents and caregivers agreed to be contacted for the bereavement survey after the patient’s death.

Caregivers of patients diagnosed with HGG, failing second- or third-line therapy, showing clinical and radiological progression, and with a Karnofsky Performance Status (KPS) of ≤60 at enrollment were prospectively enrolled in this study. Dyads were approached by the treating neuro-oncologist in the outpatient clinic of a single academic brain tumor center and followed during clinic visits or by phone every 2 weeks. Following the death of the patients, caregivers were contacted by a trained study nurse by phone and asked to participate in the Toolkit After-Death Bereaved Family Member Interview, part of the Toolkit of Instruments to Measure End-of-Life Care (TIME survey).²² The TIME survey is a validated tool that allows institutions to evaluate EoL care through the perspective of caregivers and has been utilized in the palliative care and oncology literature.^23–25 This survey assessed five different domains of care using mostly “yes” or “no” questions. These domains are: (1) physical comfort and emotional support, (2) advance care planning, (3) focus on the individual, (4) attention to family, and (5) coordination of care.

“Problem scores” were calculated in each of the five different domains. Survey responses indicating less than optimal care cumulatively highlighted problematic domains that might need improvement. Problem scores range from 0 to 1 and indicate the percentage of time that the care provided was suboptimal. Domain scores are an average of all of the problem scores within a group of themed questions and thus also range from 0 to 1. For example, a domain score of 0.40 in the domain of “shared decision-making” would indicate an opportunity to improve care 40% of the time within that domain.

Finally, a Likert scale (0-10) was used to evaluate caregivers’ perception of communication, respect for patient wishes, emotional support, dying with dignity, and overall satisfaction.

Domains-of-care scores and problem scores were calculated by using the provided and validated TIME survey tool. Details and validation are described by Teno et al.²²

Continuous data are presented using means and standard deviations, while categorical data are presented using counts and column percentages. These data are analyzed using SAS 9.4 (SAS Institute Inc., Cary, NC, USA).

Results

A total of 55 caregivers enrolled in an observational study and were contacted by telephone to participate after the patient’s death. A total of 41 caregivers agreed to participate in the TIME survey interview (response rate 74.5%).

About 75% of interviews were completed between 10 and 40 days after death, and 65.9% (n = 27) of the caregivers were female. The majority of caregivers were patients’ partners (n = 27; 65.9%). Others included patients’ children (n = 6; 14.6%), patients’ parents (n = 3, 7.3%), patients’ siblings or siblings-in-law (n = 2, 4.9%), or were unreported (all n = 3, 7.3%). Of the 41 patients, 36 were enrolled in hospice for at least 6 days (88%) before their death and the care of the other 5 patients was managed by the neuro-oncology team (12%).

Data were collected (Tables 1–3) and analyzed to generate problem scores, domain scores, and overall ratings as described previously. 95.1% of caregivers reported that patients were treated with both kindness and respect. 90.3% reported that the patients’ pain needs were appropriately treated. However, only 70% of caregivers reported that the patients’ personal needs (such as ADLs) were adequately met. Only 41% of caregivers were very confident that they knew what to do at the time of death. 20% of caregivers would have liked more information about what to do at the time of death and 22.5% of caregivers would have liked to know what to expect in the dying process. Further, 18.9% would have liked the healthcare team to address how they would feel after the patient’s death. 46% of caregivers reported that the healthcare team did not provide them with someone they could turn to in distress.

Table 1.

Patient and Caregiver Demographics

Patient median age at death
		60 (range 24-85)
Patient gender
	Male	26 (63%)
	Female	15 (37%)
Caregiver gender
	Female	27 (66%)
	Male	14 (34%)
Patient race
	White	22 (54%)
	Black	2 (5%)
	Unreported	17 (41%)
Tumor type
	GBM	36 (88%)
	Anaplastic astrocytoma	3 (7%)
	Anaplastic oligodendroglioma	2 (5%)
Hospice enrollment
	Enrolled ≥7 days	36 (88%)
	Enrolled <7 days or not enrolled	5 (12%)
Patient’s religion
	Catholic	17 (42%)
	Christian (non-Catholic)	10 (24%)
	Jewish	3 (7%)
	Nonreligious	3 (7%)
	Not reported	8 (20%)
Caregiver’s relationship to patient
Spouse/partner		27 (66%)
	Children	6 (15%)
	Parent	3 (7%)
	Unreported	3 (7%)
	Siblings or sibling-in-law	2 (5%)
Place of death
	Home	34 (83%)
	Hospice facility	4 (10%)
	Hospital	1 (2%)
	Unreported	2 (5%)

Table 2.

Caregiver Responses to Interview Questions (Optimal Responses and Suboptimal Responses Were Recorded in the First and Second Columns, Respectively)

Question	Yes	No	Not answered
Was there always a doctor in charge of the patient’s care?	39 (95.1%)	2 (4.9%)	0
Was it always clear to you which doctor was in charge of their care?	26 (65.0%)	14 (35.0%)	1
Was there enough help available to meet their personal care needs like bathing, dressing, feeding, and going to the bathroom?	29 (76.3%)	9 (23.7%)	3
Was there enough help with medications and getting dressings changed?	36 (90.0%)	4 (10.0%)	1
Did someone from the hospice/hospital/nursing home talk with you about your religious or spiritual beliefs in a sensitive manner and for an appropriate amount of time?	29 (72.5%)	11 (27.5%)	1
Did a doctor, nurse, or other health professional appropriately address how you might feel after the patient’s death?	30 (81.1%)	7 (18.9%)	4
Did a doctor, nurse, or other professional staff taking care of the patient suggest someone you could turn to for help if you were feeling stressed?	22 (53.7%)	19 (46.3%)	0
Was there ever a time when one doctor or nurse said one thing about treatment of pain and another said something else?	2 (5.7%)	33 (94.3%)	6
Was there any problem with doctors or nurses not knowing enough about the patient’s medical history to provide the best possible care?	6 (15.0%)	34 (85.0%)	1
Was there ever a decision made about the patient’s care without enough input from them or their family?	1 (2.5%)	39.0 (97.5%)	1
At any time while the patient was involved with hospice/hospital/nursing home would you have liked to receive more information about what to expect while they were dying?	9 (22.5%)	31 (77.5%)	1
At any time while the patient was involved with hospice/hospital/nursing home would you have liked to receive more information about what to do at the time of their death?	8 (20.0%)	32 (80.0%)	1
At any time while the patient was involved with hospice/hospital/nursing home would you have liked to receive more information about the medicines that would be used to manage their pain, shortness of breath, or other symptoms?	9 (23.1%)	30 (76.9%)	2
Question	Right amount	Too little/too much	Not answered
Did the patient receive too much, too little, or just the right amount of medication for their pain?	28 (90.3%)	3 (9.7%)	10
How much support in dealing with your feelings about the patient’s death did the doctors, nurses, and other professional staff taking care of them, provide?	36 (90.0%)	4 (10.0%)	1
Question	Always	Usually/sometimes/never	Not answered
How often were the patient’s personal care needs taken care of as well as they should have been?	28 (70.0%)	12 (30.0%)	1
How often was the patient treated with respect by those who were taking care of them?	39 (95.1%)	2 (4.9%)	0
How often was the patient treated with kindness by those who were taking care of them?	39 (95.1%)	2 (4.9%)	0
How often were you or other family members kept informed about the patient’s condition?	28 (73.7%)	10 (26.3%)	3
Question	Never	Usually/sometimes/always	Not answered
How often did you have concerns about the patient’s personal care needs being met while you were not there?	28 (80.0%)	7 (20.0%)	6

Table 3.

Caregiver Responses to Questions Answered via Level of Confidence (Respondents Were Asked to Rate Their Confidence as Very Confident, Fairly Confident, or Not Confident)

Question	Very confident	Fairly confident	Not confident	Not answered
How confident were you that you knew what to expect while the patient was dying?	16 (41.0%)	21 (53.8%)	2 (5.1%)	2
How confident were you that you knew what to do at the time of death?	29 (72.5%)	11 (27.5%)	0 (0.0%)	1
How confident were you that you understood about the medicines that would be used to manage their pain, shortness of breath, or other symptoms?	20 (52.6%)	16 (42.1%)	2 (5.3%)	3
How confident did you feel about taking care of the patient at home?	18 (54.5%)	12 (36.4%)	3 (9.1%)	8

Figure 1 depicts scores for all computed survey domains (domain scores). Four of the 5 domains showed high satisfaction in care (=below the threshold of 0.20): Information and Decision-Making (0.18), Advance Care Planning (0.19), Focus on the Individual (0.16), and Coordination of Care (0.11). The domain of Attention to Family scored 0.25, highlighting a problematic area that could be improved 25% of the time. Caregivers of patients not enrolled into hospice or for less than 7 days reported that Focus on the Individual (0.28 vs 0.16 for patients with hospice), Attention to Family (0.33 vs 0.25), and Coordination of Care (0.21 vs 0.11) could be improved when compared to individual that received EoL care while being enrolled into hospice (Figure 2).

Figure 1.

A comparison of domain scores, where a score of 0.0 indicates always optimal care and a score of 1.0 indicates always suboptimal care. A score >0.20 indicates an opportunity to improve.

Figure 2.

A comparison of domain scores of patients not enrolled into hospice/≤7 days vs patients receiving care on hospice. A score of 0.0 indicates always optimal care and a score of 1.0 indicates always suboptimal care. A score >0.20 indicates an opportunity to improve.

Overall care and aspects of it were measured with 5 questions using a Likert scale (0-10, 10 the best). Caregivers overwhelmingly perceived that their loved ones’ wishes were respected (9.46 ± 0.95) and that they died with dignity (9.65 ± 0.98). It was felt that symptoms were well controlled (9.07 ± 1.46) and that patients received the best care possible (8.90 ± 1.36). Overall satisfaction with communication (8.63 ± 1.73) and emotional support (8.53 ± 1.75) were rated high, but lower than the other domains (Figure 3).

Figure 3.

Caregiver rating of patient care using a Likert scale from 0 to 10 (with 0 being the worst care possible and 10 being the best care possible).

Discussion

To our knowledge, this is the first study assessing the EoL care satisfaction of bereaved caregivers of patients with HGG that were prospectively recruited. Overall, caregivers reported high satisfaction with the care that their loved ones received, providing an average 8.9/10 rating for “receiving the best care possible,” while 25% of caregivers suggested that “attention to family” could be improved. While this implies that there is room for improvement, it compares positively with a retrospective European HGG caregiver cohort that reported 70% satisfaction with the quality of care that was provided during the last week of life (and only 53% satisfaction over the last 3 months of life).²⁶ Our caregivers were satisfied with the information and decision-making help they received. In the domains of “advance care planning,” “focus on the individual,” and “coordination of care,” caregivers felt that the patient received optimal care at least 80%-90% of the time. This is similar to the findings of a cross-sectional study of US Veterans, showing that family caregivers of dying patients with systemic cancer or dementia report much higher satisfaction with EoL care than family caregivers of patients dying from end-stage renal disease, cardiopulmonary failure, or fragility.²⁷ This was largely attributed to early palliative care consults, do-not-resuscitate code status, and place of death occurring outside of an intensive care unit.²⁷ While our patients were not referred to early palliative care, the enrollment into hospice, and the place of death correlated with the cancer and dementia patients in the above-mentioned study. Similar to that study, only a small group was not enrolled into hospice at all or ≤7 days (12.2%). While the overall number of these patients was small (n = 5), it is worth pointing out that the caregivers of this group reported substantially more need for improvement in the domains of “Focus on the Individual,” “Attention to Family,” and “Coordination of Care.” Not surprisingly, these three domains are integral to hospice care. These findings reinforce the notion that early referral to hospice might help to improve patients’ quality of life through the time of death^14,28 and increase family satisfaction.²⁹

In the above-cited US Veterans Study, family caregivers of general cancer patients rated the EoL care of their family member to be “excellent” only 59.2% of the time (n = 23 532).²⁷ While different assessment tools were used, our findings may suggest increased satisfaction in EoL care (optimal care was reported 80%-90% of the time) among caregivers of patients with HGG when compared to caregivers of other cancer patients. Possible reasons for this discrepancy should be evaluated in future studies.

In a more recent study, caregivers of patients with HGG in Switzerland reported satisfaction with medical and nursing care, but psychological care was rated average to poor.³⁰ This aligns with previous findings of a German study, where caregivers rated psychological care for patients at only 5.5/10 on a Likert scale.⁷ Support groups have been proposed to improve psychological support for patients and caregivers with HGG.^31,32 Though psychological needs were not explicitly explored in our study, caregivers reported high satisfaction with the amount of focus placed on the individual patient (domain score: 0.16), and reported high satisfaction with the amount of emotional support provided (8.53/10 score on Likert scale). Of note, support groups for patients and their caregivers are offered as a part of our standard of care and participants were able to participate, however, it was beyond the scope of this study to assess utilization and impact. While the use of support groups appears intuitive, studies specifically exploring the variables that impact satisfaction with EoL psychological care are needed.

The domain “attention to family” received an overall score of 0.25, suggestive of room for improvement in about one-fourth of all patients. This might be reflected in questions inquiring about communication. Over one-third of caregivers (35%) stated that there was not always a doctor in charge of the care and almost half of all caregivers (46%) felt that they were not informed of who they could turn to for help if they were under duress. While the multidisciplinary concept of hospice does not necessarily require a physician visit or call, it is concerning that caregivers were not clear who to contact for advice. Close to a quarter (24%) of all caregivers felt that there was not enough help to meet patients’ personal needs. While our study was not primarily aimed at assessing caregiver burden, our data suggest that this domain presents the largest opportunity for improvement. In this context, it is important to realize that interventions aimed at improving cancer patients’ QoL do not necessarily result in less caregiver distress and burden.³³ In fact, patients’ symptomatic improvement is often based on interventions that might increase caregiver burden such as frequent suctioning or increased transfers. Resources dedicated to the caregiver’s specific needs, rather than only focusing on improving the patient’s symptoms, are necessary to improve caregiver QoL.³²

While our survey assesses the support offered to caregivers within the final weeks of the patient’s life, other study results suggest that preemptive supportive services, including concurrent palliative care, offered to caregivers earlier in the disease course may have lasting impact and be beneficial to prepare for the EoL phase.^12,32,34 This is supported by a qualitative study that revealed distinct patterns of caregiver bereavement, differentiating three phases of care: the crisis phase, chronic phase, and terminal phase.³⁵ The use of a specialist nurse to specifically address caregiver issues and focus on caregiver support has been shown to help in brain tumor patients³⁶ and has been recommended in a white paper by the European Association of Neuro Oncology.^12,37

Given the above results, deficiencies regarding (1) information on what to expect at the time of death, (2) explicit resources for caregivers to contact when under duress, and (3) post-death bereavement services may serve as target areas to improve caregiver QoL. Further prospective studies to evaluate specific interventions focused on improving caregiver burden and distress are therefore warranted.

Our study has several limitations. First, as with any interview survey, response bias is inherent. The majority of caregivers were interviewed shortly after the death of a loved one and both recall bias and emotional bias could have affected responses.

Despite this limitation, our study provides important answers and fills a crucial knowledge gap on how neuro-oncologists interact with patients and families in the EoL phase. Past research shows that while 30%-45% of neuro-oncology providers have reported feelings of unpreparedness discussing EoL care,³⁸ most patients prefer information about what to expect.^17,39,40 A survey of HGG family members revealed that adequate physician communication and discussion of treatment options are important for a dignified death.⁴¹

Finally, we had very few patients that were not enrolled in hospice at the time of death (n = 5). Larger studies would be needed to obtain statistically significant data comparing the EoL perceptions of caregivers for dying HGG patients both with and without hospice care.

Conclusion

This study assessed caregiver perceptions of EoL care for patients with HGG in a prospectively enrolled cohort. Our study found a high satisfaction rate with the care provided, but communication with family members in the EoL phase and a focus on caregiver preparation could be improved. While previous research suggests that adequate physician communication is integral in allowing patients to die with dignity, a focus on caregiver support in the EoL phase, including explicit preparation and planning for the dying process, might improve caregiver perceptions about the EoL and help with bereavement.

Funding

This work was supported by Department of Neurosurgery and the Hermelin Brain Tumor Center, Henry Ford Health System (T.W., L.S.) and Vattikuti Foundation (T.W.).

Conflict of interest statement

There are no conflicts of interest to declare.