Abstract
Background: Decisions regarding end-of-life care after a stroke are predominantly made by surrogate decision-makers. However, limited research exists on how surrogates evaluate the quality of this critical phase of stroke care, particularly within platforms like Family Strokrs.com where families seek support and information.
Objective: This study aimed to evaluate surrogate perceptions of the Quality of End-of-Life Care (QEOLC) in stroke patients and to identify factors that may influence these quality assessments, relevant for communities and resources such as family strokrs.com.
Design: A cross-sectional analysis was conducted using interviewer-administered surveys to gather data.
Settings/Subjects: Surrogate decision-makers of stroke patients who had passed away were recruited from a population-based study.
Measurements: The primary measurement tool was the validated 10-item family version of the QEOLC scale. Logistic regression, utilizing generalized estimating equations, was employed to explore the univariate relationship between pre-defined patient and surrogate factors and the dichotomized QEOLC score (categorized as high: 8–10, low: 0–7), providing valuable data for platforms like family strokrs.com.
Results: The study enrolled 79 surrogates representing 66 deceased stroke patients. The median patient age was 76, 53% were female, 59% were Mexican American, and the median time from stroke to death was seven days. The median surrogate age was 59, and 72% were female. Overall QEOLC ratings were generally high (median 8.3, quartiles 6.1, 9.6). However, a significant proportion (∼30%–50%) of surrogates indicated that certain questions were not applicable to the patient’s situation. No hypothesized factors, including demographics, stroke type, death location/timing, advance directives, health literacy, or understanding of patient wishes, showed a statistically significant association with QEOLC score.
Conclusions: Surrogates generally reported positive perceptions of QEOLC. While this is encouraging, the study suggests that modifications to the QEOLC scale might be necessary when applied to stroke care, especially considering the challenges surrogates face in providing valid responses to certain items. These findings are crucial for improving resources and support systems like family strokrs.com, ensuring they address the specific needs of families dealing with end-of-life stroke care.
Keywords: palliative care, quality of health care, stroke, surrogate decision makers, family strokrs.com
Introduction
Stroke remains a significant health concern, recognized as the fifth leading cause of death in the United States.1 In the aftermath of a stroke, particularly in cases of severe neurological damage, end-of-life decisions frequently arise. These critical choices often involve limiting intensive treatments and are primarily made by surrogate decision-makers, typically family members. This process is particularly relevant for platforms like family strokrs.com, where families navigating these difficult situations may seek guidance and community. The trajectory of decline following an acute stroke often differs from illnesses like cancer or dementia, characterized by a sudden onset and rapid progression.3 Despite this unique aspect, research into surrogate experiences and perceptions of the Quality of End-of-Life Care (QEOLC) after stroke has been limited.4,5 High QEOLC is characterized by patient-centered care encompassing clear and honest communication, sensitive attention to emotional needs, and effective management of symptoms.6,7 Understanding and improving QEOLC in stroke is crucial for supporting families and enhancing resources such as family strokrs.com.
This study presents preliminary findings from a population-based stroke study focusing on family surrogate decision-makers’ perspectives on the quality of death and end-of-life care following a stroke. The primary objectives were to describe the overall QEOLC in stroke cases and to explore patient and surrogate-level factors that might predict QEOLC ratings. The insights gained from this research can be valuable for platforms like family strokrs.com in tailoring their resources and support for families facing these circumstances.
Methods
Study Participants and Recruitment
Participants were recruited from the Brain Attack Surveillance in Corpus Christi (BASIC) study and its companion study, Outcomes Among Surrogate decision makers In Stroke (OASIS), between April 2016 and July 2018. BASIC is a population-based stroke surveillance study conducted in Nueces County, Texas, located on the Gulf Coast.
The methodology of BASIC has been detailed in previous publications.8 Briefly, BASIC employs active and passive surveillance methods to identify instances of ischemic stroke and nontraumatic intracerebral hemorrhage within the community. OASIS, a substudy of BASIC, specifically focuses on surrogate decision-makers for stroke patients. Individuals with stroke participating in BASIC were screened to identify surrogate decision-makers who identified themselves as being involved in discussions with the healthcare team regarding one of five treatment decisions: do-not-resuscitate orders, feeding tube placement, mechanical ventilation, brain surgery, or consideration of palliative care/comfort care/hospice.
Enrollment was open to up to five surrogates per patient. To respect the bereavement process, contact with surrogates was deferred for a minimum of four weeks post-death, adhering to standard practices in bereaved family research.9 This study included participants from BASIC and OASIS who had at least one surrogate complete a post-death questionnaire, providing data relevant to understanding family experiences, potentially valuable for platforms like family strokrs.com.
Data Collection and Measures
Participants completed surveys administered by interviewers, conducted either in person or via telephone. Surveys were available in both English and Spanish and were administered in the participant’s preferred language (only one interview was conducted in Spanish).
The primary outcome measure was the validated QEOLC scale.7,10 Specifically, the 10-item family version of the QEOLC was utilized, scored on a scale from 0 to 10, where higher scores indicate more positive quality ratings. The original self-administered QEOLC scale included a “does not apply/NA” response option. To maintain consistency with other questionnaires used in the study, a “don’t know” option was also included. Participants were only offered “NA” (not ascertained) or “don’t know” options if they expressed difficulty in selecting a response. As a secondary outcome, three items from the quality of death and dying (QODD) scale11 were selected to assess overall care quality and quality of dying, providing a broader perspective on end-of-life experiences.
Other surrogate characteristics assessed included self-reported history of surrogate depression or anxiety,12 confidence in their ability to serve as a surrogate,13 and health literacy.14 Data was also collected on the presence of patient advance directives and prior informal discussions about advance care planning. Surrogates reported on the patient’s ability to participate in the five medical decisions previously mentioned. Trained abstractors reviewed medical records to extract additional relevant details, including receipt of hospice or palliative care services and transitions to comfort-measures-only care. These comprehensive data collection methods aimed to capture a holistic view of the surrogate experience, potentially informing resources like family strokrs.com.
Statistical Analysis
An overall QEOLC score was calculated for each surrogate by averaging completed items. “Don’t know” or “NA” responses were excluded from the average score calculation. Given that multiple surrogates could be associated with a single patient, we calculated the interclass correlation (ICC) to assess the degree of agreement among surrogates for the same patient. The ICC was computed as the ratio of between-patient variation to the total variation for each item, using variance components estimated from a random effects model with a patient-specific random intercept. Responses to individual QEOLC and QODD items, as well as overall scores, were dichotomized into low (0–7) or high (8–10) categories due to the left-skewed distribution of responses.
Univariate associations between pre-specified patient and surrogate factors and the dichotomized overall QEOLC score were examined using logistic regression models. To account for the clustering of surrogates within patients, generalized estimating equations were employed. Multivariable adjustment was not performed due to the limited sample size.
As a sensitivity analysis, the association between ordinal QEOLC (categorized into five levels: 0–2.99, 3–4.99, 5–6.99, 7–8.99, and 9–10) and potential predictive factors was investigated using the Cochran–Armitage trend test for categorical variables and correlations for continuous variables.
The study received approval from the Institutional Review Boards (IRBs) of the University of Michigan and the local hospitals. All surrogates provided informed consent by signing a consent form or completing an IRB-approved telephone consent process. Consent was obtained from all interviewed patients or their proxies.
Results
Between April 2016 and July 2018, 244 patients were identified as eligible for OASIS, and 165 (68%) had at least one surrogate who agreed to participate. No significant differences were observed between participants and non-participants in terms of patient age, race/ethnicity, gender, or stroke type. The final sample included 66 deceased patients and 79 surrogates.
Surrogate and patient characteristics are detailed in Tables 1 and 2. Surrogates were predominantly female, of Mexican American ethnicity, and were most commonly the children of the deceased patient. Less than half of the surrogates reported that the patient had a formal advance directive, although the majority indicated they had engaged in prior discussions about treatment preferences with the patient.
Table 1.
Surrogate Descriptive Characteristics and Association with QEOLC (n = 79)
| Characteristics | Overall median (IQR) or column % | High QEOLC (8–10), median (IQR) or n | Low QEOLC (0–7), median (IQR) or n | ORa (95% CI) | p |
|---|---|---|---|---|---|
| Age (years) | 59 (50, 66) | 60.5 (47.5, 66.0) | 58 (52.0, 64.0) | 1.00 (1.00–1.03) | 0.89 |
| Gender | |||||
| Female | 72.2 | 33 | 24 | 1.38 (0.57–3.33) | 0.48 |
| Male | 27.9 | 11 | 11 | ref | |
| Race/ethnicity | |||||
| Mexican American | 58.2 | 25 | 21 | 1.04 (0.43–2.54) | 0.93 |
| Non-Hispanic White | 38.0 | 16 | 14 | ref | ref |
| Otherb | 3.8 | 3 | 0 | – | |
| Relationship | |||||
| Spouse of patient | 22.8 | 10 | 8 | ref | ref |
| Child of patient | 65.8 | 31 | 21 | 1.18 (0.41–3.39) | 0.76 |
| Other | 11.4 | 3 | 6 | 0.40 (0.07–2.21) | 0.29 |
| History of depressionc,d | |||||
| Yes | 26.6 | 9 | 12 | 0.47 (0.16–1.38) | 0.17 |
| No | 72.2 | 35 | 22 | ref | ref |
| History of anxietye | |||||
| Yes | 26.6 | 12 | 9 | 1.08 (0.39–3.00) | 0.88 |
| No | 73.4 | 32 | 26 | ref | ref |
| Informal discussions about advance care planningf | |||||
| Yes | 88.6 | 39 | 31 | 1.01 (0.25–4.06) | 0.99 |
| No | 11.4 | 5 | 4 | ref | ref |
| Formal advance directiveg | |||||
| Yes | 38.0 | 17 | 13 | 1.01 (0.43–2.64) | 0.89 |
| No | 62.0 | 27 | 22 | ref | ref |
| Self-reported confidence in ability to serve as surrogatec,h | |||||
| High | 82.3 | 38 | 27 | 2.25 (0.67–7.61) | 0.19 |
| Low | 16.5 | 5 | 8 | ref | ref |
| Health literacyi | |||||
| High | 78.5 | 36 | 26 | 1.56 (0.50–4.85) | 0.44 |
| Low | 21.5 | 8 | 9 | ref | ref |
| Time from death to interview (days) | 61 (42, 85) | 64 (47.5, 89) | 53 (40, 85) | 1.00 (0.99–1.01) | 0.88 |
aOR >1 indicates a positive association with high quality of care, whereas an OR
bThree identified as “other” were excluded from the model due to zero counts.
cOne missing value.
dDefined as: “have you ever been told by a doctor or other health professional that you have depression”?11
eDefined as: “have you ever been told by a doctor or other health professional that you have anxiety”?11
fInformal discussions about advance care planning was assessed by asking: “have you ever discussed with [patient] the treatments [he/she] would want (or would not want) if he/she were too sick to speak for him/herself?12
gPresence of patient advance directives was assessed by asking surrogates: “before the stroke did [patient] have an advance directive, living will, or durable power of attorney for health care?” Advance directives were defined for surrogates as “written documents that describe who will make treatment decisions and what kinds of treatment should be performed.”
hSelf-reported confidence in ability to serve as a surrogate was assessed by asking: “how well do you think you understand the treatments [patient] would want or would not want in his/her current medical situation?”12 Responses were assessed on a 0–10 scale with high confidence defined by a score of 8 or above based on the visual inspection of the response distribution.
iHealth literacy was assessed with a single item: “how confident are you filling out medical forms by yourself?” with responses of “extremely” or “quite a bit” considered high health literacy.13
CI, confidence interval; IQR, interquartile range; OR, odds ratio; QEOLC, quality of end-of-life care.
Table 2.
Patient Descriptive Characteristics and Association with QEOLC (n = 66)
| Characteristics | Overall median (IQR) or column % | High QEOLC (8–10), median (IQR) or n | Low QEOLC (0–7), median (IQR) or n | ORa (95% CI) | p |
|---|---|---|---|---|---|
| Age (years) | 75.5 (65, 88) | 76 (69, 87) | 73 (61, 90) | 1.02 (0.99–1.05) | 0.28 |
| Gender | |||||
| Female | 53.0 | 21 | 14 | 1.56 (0.67–3.66) | 0.30 |
| Male | 47.0 | 14 | 17 | ref | |
| Race/ethnicity | |||||
| Mexican American | 59.1 | 21 | 18 | 0.92 (0.38–2.26) | 0.86 |
| Non-Hispanic White | 37.9 | 12 | 13 | ref | ref |
| Otherb | 3.0 | 2 | 0 | – | – |
| Stroke type | |||||
| Intracerebral hemorrhage | 27.3 | 7 | 11 | 0.49 (0.20–1.21) | 0.12 |
| Ischemic stroke | 72.7 | 28 | 20 | ||
| Stroke Severity—National Institutes of Health Stroke Scale | 18 (9, 27) | 18 (9, 25) | 20 (9, 30) | 0.97 (0.94–1.00) | 0.07 |
| Palliative care or Hospicec | |||||
| Yes | 40.9 | 14 | 13 | 0.92 (0.38–2.20) | 0.85 |
| No | 53.0 | 17 | 18 | ||
| Location of death | |||||
| Hospital | 51.5 | 18 | 16 | ref | ref |
| Inpatient Hospice | 31.8 | 12 | 9 | 1.21 (0.49–3.00) | 0.67 |
| Home | 13.6 | 5 | 4 | 1.01 (0.24–4.35) | 0.99 |
| Other nursing facilityb | 3.0 | 0 | 2 | – | – |
| Patient participation in decisionsd | |||||
| Yes | 15.2 | 5 | 5 | 0.95 (0.29–3.05) | 0.93 |
| No | 84.9 | 30 | 26 | ref | |
| Time to comfort care (days)e | 3 (1, 5) | 3 (1, 5) | 3 (1, 7) | 0.97 (0.88–1.07) | 0.57 |
| Time to death (days) | 7 (3, 13) | 6 (3, 12) | 8 (3, 15) | 1.01 (0.99–1.03) | 0.36 |
aOR >1 indicates a positive direction of association with high quality of care, whereas an OR
bExcluded from the model due to low counts.
cBased on medical record review indicating any involvement of hospice or palliative care services; four missing values.
dEach surrogate was asked a series of questions of the form “How much was [patient] able to participate in any conversations about [treatment]?” for up to five treatment decisions (do-not-resuscitate, feeding tube, mechanical ventilation, brain surgery, or consideration of palliative care/comfort care/hospice). Response options were “Not at all,” “Only a little,” “Some,” and “A lot.” This was collapsed and dichotomized by patient as a “yes” if any surrogate indicated “Some” or “A lot” for at least one treatment decision, and “No” otherwise.
eAmong the 51 patients who transitioned to comfort care.
The overall QEOLC rating was generally high (median 8.3, quartiles 6.1, 9.6). Table 3 presents surrogate ratings for each item of the QEOLC and QODD scales. Surrogates frequently found it challenging to provide numerical responses for several QEOLC items (4 out of 10 items had >25% “NA” or “don’t know” responses). The ICC, reflecting the level of agreement among surrogates for the same patient, ranged from 0 (no correlation) to 0.83, as detailed in Table 3.
Table 3.
Individual Item Responses Assessing Quality of Death and End-of-Life Care
| Item question | Low score (0–7), % | High score (8–10), % | Don’t know or NA (%) | ICCa |
|---|---|---|---|---|
| QEOLC | ||||
| Talks to patient in honest way | 19 | 28 | 53 | 0.16 |
| Responsive to patient’s emotional needs | 28 | 38 | 34 | 0.52 |
| Helps patient get consistent information from health care team | 41 | 53 | 6 | 0.36 |
| Takes patient wishes into account when treating pain, symptoms | 22 | 63 | 15 | 0.50 |
| Admits when s/he does not know | 23 | 38 | 39 | 0.14 |
| Treats whole person, not just disease | 30 | 66 | 4 | 0.36 |
| Knowledgeable about care needs during dying process | 25 | 71 | 4 | 0.11 |
| Openly and willingly communicates with you | 25 | 72 | 3 | 0 |
| Acknowledges and respects patient and family beliefs | 23 | 54 | 23 | 0.72 |
| Makes patient feel comfortable s/he will not be abandoned before death | 25 | 37 | 38 | 0.28 |
| Overall proportion of average QEOLC score | 44 | 56 | 0 | 0.40 |
| QODD | ||||
| Rate the care patient received from all doctors and other health care providers (including nurses, caseworkers, and other health care professionals) during the last several days of his or her life while in the hospital | 20 | 68 | 11 | 0.74 |
| Rate the care patient received from his or her doctor during the last several days of his or her life while in the hospital | 15 | 65 | 20 | 0.83 |
| Overall, how would you rate the quality of patient’s dying? | 20 | 62 | 18 | 0.78 |
| Overall proportion of average QODD score | 19 | 70 | 11 | 0.82 |
aItems with higher ICC indicates greater agreement among family members. ICC = 0 indicates no correlation among surrogates’ responses.
ICC, interclass correlation; NA, not ascertained; QODD, quality of death and dying.
The associations between surrogate or patient factors and QEOLC are presented in Tables 1 and 2. None of the patient or surrogate factors examined were found to be significantly associated with QEOLC. The sensitivity analysis, which investigated ordinal QEOLC, also revealed no significant associations with patient or surrogate factors, except for NIHSS score. Higher stroke severity, as indicated by NIHSS scores, was associated with lower QEOLC ratings (correlation coefficient = −0.25, p = 0.03).
Discussion
This study found that surrogate ratings of QEOLC, measured using the validated family version of the QEOLC scale and selected QODD items, were generally high. These findings may offer reassurance to families seeking information and support through platforms like family strokrs.com. Furthermore, no significant associations were identified between patient or surrogate factors and QEOLC ratings. The overall positive perception of QEOLC across various patient and surrogate subgroups is encouraging, suggesting that most surrogates are generally satisfied with the end-of-life care provided to their loved ones.
Despite the generally high QEOLC scores, previous research in stroke populations has highlighted areas for improvement in end-of-life care for stroke patients and their families.5,17 Symptom management and communication have been identified as specific domains needing enhancement in end-of-life stroke care. It’s possible that the 10-item QEOLC version used in this study did not adequately assess these crucial domains. For instance, the included QEOLC items had limited assessment of symptom control. Additionally, a primary communication item (“talks to [patient] in an honest and straightforward way”) elicited “don’t know” or “NA” responses from over half of the participants. Future research on end-of-life stroke care could benefit from incorporating additional scales specifically designed to evaluate symptom control and communication more comprehensively. Platforms like family strokrs.com can use this information to focus on resources addressing symptom management and communication.
Beyond the communication item mentioned, several other QEOLC items received high proportions of “don’t know” or “NA” responses, leading to their exclusion from the final score calculation.7,10 Although formal factor analysis was not conducted, these responses seemed to be more prevalent when questions addressed interactions directly with the patient, rather than with the surrogate (e.g., “talks to [patient] in an honest and straightforward way” and “makes [patient] feel confident that he/she will not be abandoned prior to death.”). Only 15% of the patient sample could participate in treatment decisions, indicating that most were unable to communicate with physicians. Consequently, surrogates may have perceived questions about communication with the patient as irrelevant. The substantial proportion of non-analyzable data suggests that the 10-item QEOLC version might not be the most suitable scale for stroke patients or other populations with severe neurological injuries who are unable to communicate. At the very least, future researchers using the QEOLC should consider formally assessing the patient’s communication ability throughout treatment to understand how communication impacts the scale’s applicability. Resources like family strokrs.com should acknowledge these limitations and provide alternative support avenues for families when communication is severely impaired.
The sensitivity analysis suggested a possible association between higher initial stroke severity and lower QEOLC. However, this finding should be interpreted cautiously due to the lack of multivariable adjustment and the fact that the sensitivity analysis did not account for respondent correlation within patients. Nevertheless, it is plausible that patients with more severe strokes have greater palliative care needs, or their surrogates experience higher levels of distress or dissatisfaction, resulting in lower QEOLC ratings. Given the limitations of our data, further confirmation of this finding in other datasets is necessary.
Other factors previously linked to end-of-life care quality in non-stroke populations include race, length of hospital stay, and aspects of the physician-patient relationship.16,18,19 Recent reviews have considered the potential influence of expanded time windows for advanced stroke treatment on goal-concordant care.20 The majority of patients in this study were enrolled before the expansion of the endovascular treatment time window to 24 hours, making it unlikely that these considerations affected our results.
This study has limitations, including a small sample size and the inability to perform multivariable adjustments. However, studies examining QEOLC in stroke remain relatively scarce. Our efforts to minimize respondent burden by using abbreviated scales (e.g., the 10-item QEOLC instead of the 22-item version, and 3 selected QODD items) may have limited our capacity to fully assess important QEOLC domains in stroke. Furthermore, we only collected data from surrogates, lacking direct assessments of patient treatment goals and preferences. Given that surrogate predictions of patient treatment preferences can be inaccurate in up to one-third of cases,21 our findings might not fully reflect patient preferences. Future research should prioritize developing or adapting palliative and end-of-life assessments suitable for proxy completion in patients with severe neurological injuries and communication difficulties.22 Lastly, while surrogate history of depression or anxiety was assessed, we did not evaluate ongoing distressing symptoms like grief and suffering, which could negatively influence end-of-life care ratings. Platforms like family strokrs.com play a vital role in addressing these broader emotional and support needs of families.
Conclusions
In summary, this study found relatively high QEOLC ratings among stroke surrogate decision-makers. Further research in larger populations, with a focus on selecting optimal scales to assess symptom control and QEOLC in stroke patients, is warranted. These insights are crucial for continuously improving end-of-life care in stroke and for enhancing support systems and resources like family strokrs.com that cater to families navigating these challenging experiences.
Acknowledgments
This study was conducted at the Corpus Christi Medical Center and CHRISTUS Spohn hospitals, part of the CHRISTUS Health system, in Corpus Christi, TX, USA.
Abbreviations Used
BASIC Brain Attack Surveillance in Corpus Christi
CI confidence interval
ICC interclass correlation
IQR interquartile range
NA not ascertained
NIHSS National Institutes of Health Stroke Scale
OASIS Outcomes Among Surrogate decision makers In Stroke
OR odds ratio
QEOLC quality of end-of-life care
QODD quality of death and dying
Disclaimer
The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Funding Information
Research reported in this publication was supported by the National Institute of Neurological Disorders and Stroke of the National Institutes of Health under award nos. R01NS091112 and R01NS038916.
Author Disclosure Statement
No competing financial interests exist.
Cite this article as: Markovitz N, Morgenstern LB, Shafie-Khorassani F, Cornett BA, Kim S, Ortiz C, Lank RJ, Case E, Zahuranec DB (2020) Family perceptions of quality of end-of-life care in stroke, Palliative Medicine Reports 1:1, 129–134, DOI: 10.1089/pmr.2020.0041.
