Palliative Home Care Based on Clinically Relevant Scientific Measures: A Cross-Sectional Study

Abstract

Introduction:

Measuring palliative care needs of patients with multiple complex illnesses and their family, is essential for providing quality clinical care. The integrated palliative care outcome scale (IPOS) is a scientifically verified measure of patients’ physical, emotional symptoms, and their palliative care communication and practical needs. The patients in this study require palliative care due to both their end-of-life heart failure (HF) and vascular dementia.

Purpose:

The purposes are to describe and compare home palliative care needs measured by the family caregivers (N = 20) and patients (N = 20) IPOS scores and to examine whether the patient IPOS total score can predict patient HF health status and caregiving burden scores.

Methods:

This study uses explanatory research design with rigorous methods for obtaining information from multiple sources. Descriptive, Cohen’s Kappa (k) statistics comparing patient and caregiver IPOS scores and regression analyses to examine the patient IPOS scores impact on patient HF health status and caregiving burden scores were used.

Results:

There was significant agreement between patient and caregiver ratings on 16 out of 17 IPOS items (k = .34 to .80). Regression analyses found that the patient IPOS total score significantly predicted patients’ HF health status (β = −.50, P < .05), and caregiving burden scores (β = .57, P < .01).

Conclusion:

Patients’ and their caregivers’ IPOS scores agreement indicates palliative care needs can be consistently identified. The patient IPOS total scores can predict patients’ HF health status and caregiving burden. These measures provide information directly applicable for health professionals guiding palliative home care.

Introduction

Heart failure (HF) and vascular dementia are serious health conditions leading to end-of-life palliative care needs.^1–3 Vascular dementia is characterized by problems with reasoning, judgment, and memory caused by impaired blood flow to the brain due to HF.⁴ HF has a progressive decline typically over 5 years until death.⁵ Family members provide the majority of home palliative care for these patients,⁶ and those family members are frequently overwhelmed,^7,8 resulting in caregiving burden.^9,10

Caregiving burden has a negative impact on family caregivers’ physical and mental health status.¹⁰ Caregivers experiencing high levels of burden often have difficulty in providing quality palliative home care, which leads to poor end-of-life issues.¹¹ Caregivers universally report being un-prepared for managing HF, dementia, and palliative home care.¹² Likewise, our previous studies with these caregivers found they lack specific information on HF disease care and on palliative home care management.^8,13,14 There is a gap of knowledge about measuring palliative home care needs for the complex end-of-life illnesses of HF and vascular dementia.^8,15

An important measure which identifies patients’ symptoms and palliative care needs is the integrated palliative care outcome scale (IPOS).¹⁶ The total IPOS score summarizes the overall patient physical and emotional symptoms, as well as communication and practical issues of palliative care.¹⁷ Multidisciplinary palliative care professionals selected and tested the 17 items on this scale based on their experiences in palliative care.¹⁸ Palliative medicine journals have reported multiple studies verifying IPOS scientific validation and responsiveness¹⁹ to HF symptoms and palliative care concerns over time.^20,21 For home caregivers and palliative care professionals, the IPOS is a valuable measure to understand the needs of the patients for whom they are caring.²²

The purposes of this study are to describe and compare home palliative care needs measured by the family caregiver and patient IPOS scores and to examine whether the patient IPOS total score can predict patient HF health status and caregiving burden scores. The patients in this study require palliative care due to both their end-of-life heart failure (HF) and their vascular dementia. The research questions for this study are:

1. What are the symptoms and palliative care needs among patients with HF and vascular dementia, rated on the patients’ IPOS and on a separate IPOS rated by their family caregivers?

2. What are the weighted Cohen’s Kappa (k) agreement scores between patient and caregiver ratings on each IPOS item?

3. Does the patient IPOS total score predict the HF patient health status and the family caregiving burden scores?

Conceptual Framework

This study is guided by the Caregiving Preparedness Model.^23–25 Caregiving preparedness is defined as possessing the information, skills, ability, and resources to provide family home caregiving.^26–28 This model verified the positive impact of caregiving preparedness on the patients’ health outcomes and caregivers’ level of burden.^14,29 Based on this model, families can be prepared for patients’ specific end-of-life symptoms related to their illnesses and their home palliative care needs.

Methods

Sample and Setting

This study used explanatory research design based on the Best Practices for Mixed Methods Research in the Health Sciences.³⁰ The study was conducted in a rural Appalachian state during September 2020 through December 2022. The University Institution Review Board approved the study (IRB No. 2006024614). Eligible patients were 55 years and older, requiring palliative care due to their medical diagnosis of HF with shortness of breath upon activity,³¹ and mild to moderate vascular dementia per the Functional Assessment Staging.³² Potential participants were identified via our Medical Center statewide integrated data repository (IDR).³³ From this IDR, 50 patients were eligible, 20 patients and their family care-givers agreed to participate and were enrolled. Nonparticipation was due to inability to reach the patients and their family caregivers (n = 10), incomplete medical records (n = 3), caregivers did not want to participate (n = 4), and no specific reasons (n = 9). Patients or family caregivers with a disability that prevented them from completing the IPOS ratings were excluded. Both patients and family caregivers were enrolled as pairs to be compared. All patients and caregivers signed consent forms to participate.

The power analysis and sample size justification were calculated using the G*Power v3.9.1.4 software. The sample size (N = 40 subjects from 20 pairs), ensured that with a significance level of 0.05, a power of 90% could be reached.³⁴ All the analyses were conducted using SPSS version 29 (IBM, Corp, 2021).

Data Collection

Patients with HF and vascular dementia and their family caregivers completed separate short surveys that included their demographic characteristics, and each patient and caregiver separately completed an IPOS scale to measure the patient’s symptoms and palliative care needs. Additionally, patients completed the Kansas City Cardiomyopathy Questionnaire (KCCQ) which measures HF states. The family caregivers completed the Zarit caregiving burden scale (ZBI-12). These combined measures took 20–30 minutes to complete. The participant responses were entered into a secure Qualtrics database and exported to SPSS version 29 for data analysis.

Measures

Demographic Variables.

Patients and caregivers completed demographic questions including their age, gender, race/ethnicity, level of education, employment, health insurance, and years of home caregiving for HF and dementia care.

Dependent (Outcome) Variables.

Patient’s HF health status was measured by the Kansas City Cardiomyopathy Questionnaire (KCCQ-12), a Likert type self-reported measure.³⁵ KCCQ-12 consists of 12 questions related to HF symptoms and their impact on daily life. The KCCQ-12 overall summary is the average of scores across the four domains of HF symptom frequency, physical limitation, quality of life, and social limitations. The score ranges from 0 to 100. Higher scores indicate better HF health status. Scores below 50 indicate poor HF health status. A recent randomized clinical trial showed that KCCQ-12 is an excellent predictor of cardiovascular outcomes and mortality.³⁶ Research found improved clinician accuracy of HF health status assessment when using the KCCQ-12 as well as improved communications with their patients.³⁷ Cronbach’s alpha reliability of the KCCQ Overall Summary HF health status score in this sample was .92.

Caregiving burden was measured by the 12-item Zarit Burden Interview, version 12 (ZBI-12).³⁸ The ZBI-12 is a Likert scale with score range from 0 = never to 4 = nearly always. The possible total scores range from 0 to 48, where a higher score indicated greater burden. A sample item question was “Do you feel stressed between caregiving and meeting other responsibilities (work/family)?” The scale has been used across HF caregiving research and in several studies involving caregivers of individuals with cognitive decline.³⁹ In this study, Cronbach’s alpha reliability of the ZBI-12 was .92.

Independent (Predictive) Variable.

The Integrated Palliative Care Outcome Scale (IPOS) is a valid and reliable measure to assess the patients’ main symptoms and concerns.¹⁹ The scientifically verified IPOS is a 17-item, 5-point Likert scale with three subscales.⁴⁰ The 10-item patient IPOS physical subscale (e.g., pain, weakness, poor mobility, etc.) ranges from 0 (not at all) to 4 (overwhelming/severe). The 4-item IPOS emotional subscale includes patient anxiety, depression, feelings at peace, and caregiver anxiety. In addition, there is a two-item IPOS subscale to measure communication issues (caregiver/patient sharing feelings, information needs). Scores range from 0 (not at all) to 4 (always). The last item on the IPOS scale is a question of whether there are practical problems related to the patient’s palliative care. Scores range from 0 (no problems) to 4 (problems not addressed). The psychometric properties of the IPOS have been verified.¹⁸ The Cronbach’s alpha reliability scores were .84 for the total IPOS scale, .80 for physical subscale, .78 for emotional subscale, and .64 for communication/practical concern subscale.

In addition, there was an open-ended question for both the patient and for the caregiver. This question was “What have been your main problems or concerns over the past 3 days?” The responses from the open-ended questions were categorized, summarized, and reported.

Data Analysis

Descriptive statistical analysis, mean (M) and standard deviation (SD) were used for demographics, predictive, and outcome variables. Prevalence of the palliative care symptoms and concerns were calculated in this sample using frequencies of patient and caregiver ratings for each IPOS item. The IPOS total and subscales scores between patients and their caregivers were calculated and compared using paired t-test analysis. To calculate agreement between patient and caregiver on IPOS individual item ratings, Cohen’s Kappa statistics were used.⁴¹ Weighted kappa scores were calculated for each of the 17 IPOS items.⁴² Generally, k scores could range from 0 to 1. The higher k scores indicate stronger agreement between patient and caregiver IPOS item ratings.

The association of independent/predictive variable (patient IPOS total scores) with dependent variables (patient HF health status and caregiving burden) (P < .05) were entered to the multiple regression analyses.⁴³ In the regression analyses, the adjusted R² identifies the proportion of variance of the dependent variable explained by the independent variables. The regression coefficient - beta (β) indicates each independent variable’s positive or negative relationship with the dependent variable. A significant F ratio with P < .05 in a regression analysis indicates a statistically significant in the sample. In addition, the responses to the IPOS open-ended question asking patients and caregivers about their most important home palliative care problems or concerns are summarized. There were 2 missing demographic responses as identified in the Table 1 legend.

Table 1.

Demographic Variables and Frequencies of Patients (N = 20) and Family Caregivers (N = 20).

Demographic Variables	Patients (N = 20)	Caregivers (N = 20)
	Freq (n, %)	Freq (n, %)
Gender
Female	9 (45.0)	16 (80.0)
Male	11 (55.0)	4 (20.0)
Ethnicityb
Non-hispanic	19 (95.0)	18 (100%)
Hispanic	1 (5.0)	0
Race (n, %)
White	19 (95.0)	19 (95.0)
Black	1 (5.0)	1 (5.0)
Marital status (n, %)
Married	8 (40.0)	14 (70.0)
Widowed, divorced, never married	12 (60.0)	6 (30.0)
Relationship of caregivers with patient
Spouse (husband/wife)		8 (40)
Daughter/son		9 (45)
Sibling or significant other		3 (15)
Education level (n, %)a
≤High school	5 (25.0)	6 (30.0)
Vocational/some college	11 (55.0)	7 (35.0)
Completed college or more	3 (15.0)	7 (35.0)
Employed (yes, no)
Yes	2 (10.0)	5 (25.0)
No	18 (90.0)	15 (75.0)
Health insurance
Medicare	14 (70.0)	8 (40.0)
Private	4 (20.0)	7 (35.0)
Medicaid	2 (10.0)	3 (15.0)
Others (e.g., military)	–	2 (10.0)
Income adequacy ratings
(1) I can’t make ends meet	2 (10.0)	2 (10.0)
(2) I have just enough; no more	7 (35.0)	5 (25.0)
(3) I have enough, a little extra sometimes	9 (45.0)	6 (30.0)
(4) I always have money left over	2 (10.0)	7 (35.0)
Number of people in the household (median)	2.00 (range 2 – 4)	2.00 (range 2 – 4)
Age in years (M, SD, and range)	79.00 (6.81)	64.95 (12.42)
	(Range 64–87)	(Range 40–80)

^aMissing patient data (n = 1);

^bMissing caregiver data (n = 2).

Results

Demographics

Of the 20 patients enrolled, 11 (55%) were male, 8 (40%) were married, and average age was 79 years (SD = 6.81). Patients had been diagnosed with HF for 9.67 years (SD = 7.28) and dementia for 4.29 years (SD = 4.82). Most of the patients were non-Hispanic white (n = 19, 95%), non-employed (n = 18, 90%). All patients had health insurance coverage (70% Medicare, 20% Private, and 10% Medicaid). Average number of people in the household was 2 (ranging from 2 to 4), but 6 (30%) of patients lived alone with their caregiver coming in to check on them. (See Table 1).

Family caregivers in this study were daughters/sons (n = 9, 45%), spouses (n = 8, 40%), and siblings or significant others (n = 3, 15%) of the patients. Caregivers were predominantly white (n = 19, 95%), and female (n = 16, 80%), with the average age of 64.95 (SD = 12.42, range 40 to 80) years. These caregivers had provided home HF caregiving for an average of 6 years (SD = 4.97) and dementia caregiving for an average of 4.29 years (SD = 4.82). One caregiver reported the duration of HF and dementia caregiving as 20 years. In terms of income adequacy, five (25%) caregivers reported “having just enough, and no more” to pay monthly bills, while 6 (30%) reported “having enough, with a little extra sometimes.” On the extreme responses of the scale, two (10%) reported “can’t make ends meet,” while seven (35%) reported “always have money left over.”

There were no significant associations between demographic variables, the IPOS total scores, the patient HF health status and caregiving burden scores. Using paired t-test statistics, there were no significant differences between patients’ and caregivers’ IPOS total and any subscale scores ratings.

Results of Research Question 1.

What are the symptoms and palliative care needs among patients with HF and vascular dementia, rated on the patients’ IPOS and on a separate IPOS rated by their family caregivers?

Table 2 reports the frequencies of symptoms and concerns reported by patients and their caregivers, highlighting commonalities reported by greater than 70% of participants. Over 70% of patients and caregivers both ranked patients’ major physical symptoms as shortness of breath, weakness, and poor mobility. Over 70% of patients and caregivers also rated all the emotional symptoms (for examples, patient anxiety, depression, feeling at peace, and family anxiety) as being high and reported challenges of sharing feelings as the most common communication problem. Notably, 70% of caregivers, compared with 60% of patients, rated patient pain as a concern, and 80% of caregivers, compared with 65% of patients, rated palliative information needs as a concern.

Table 2.

Patients (N = 20) and Their Family Caregivers (N = 20) Frequency Ratings of the Scientifically Verified IPOS Items.

IPOS Items are Listed Under the Bolded Subscales	Patient Rating	Caregiver Rating
	Freq (%)	Freq (%)
Physical symptoms subscale
Painb	60	70
Shortness of breatha	85	70
Weaknessa	90	90
Nausea	30	30
Vomiting	25	15
Poor appetite	65	65
Constipation	55	55
Sore or dry mouth	50	55
Drowsiness	65	65
Poor mobilitya	80	80
Emotional symptoms subscale
Patient anxietya	80	70
Family anxietya	85	80
Patient depressiona	70	75
Patient feeling at peacea	70	80
Communication/Practical issues subscale
Sharing feelingsa	80	80
Informationb	65	80
Practical issues	65	65

^aMajor IPOS problem/concern items rated by both patients and caregivers at frequencies ≥ 70%.

^bProblem or concerns rated only by caregivers as frequent (70% to 80%) were patient pain and information needs concern.

Results of Research Question 2.

What are the weighted Cohen’s Kappa (k) agreement scores between patient and caregiver ratings on each IPOS item?

Weighted Cohen’s Kappa (k) statistics identified significant agreement between patient and caregiving ratings on 16 out of 17 IPOS items (k = .34 to .80). The higher weighted k scores indicated stronger agreement between patients and caregivers (Table 3). One non-significant IPOS item (k = .17) between patient and caregiver was whether any practical problems resulting from patient’s illness have been addressed.

Table 3.

Weighted Cohen’s Kappa (k) Agreement^a and Asymptotic 95% CI^b Between Patient and Family Caregiver Ratings of Each IPOS Items.

IPOS Items are Listed Under the Bolded Subscales	Weighted Cohen’s Kappaa	Asymptotic 95% CIb Ranges
Physical symptoms subscale
Pain	.51	.22–.80
Shortness of breath	.49	.21–.78
Weakness	.54	.32–.77
Nausea	.44	−.01–.89
Vomiting	.46	−.14–.1.05
Poor appetite	.73	.55–.92
Constipation	.59	.31–.87
Sore or dry mouth	.62	.40–.84
Drowsiness	.66	.46–.87
Poor mobility	.45	.17–.73
Emotional symptoms subscale
Patient anxiety	.30	−.07–.67
Family anxiety	.34	.03–.66
Patient depression	.55	.22–.88
Patient feeling at peace	.39	.03–.74
Communication/practical issues subscale
Sharing feelings	.40	.08–.73
Information	.41	.07–.76
Practical issuesc	.17	−.07–.40

^aThe weighted k scores indicated agreement between patient and caregiver IPOS items. Higher k scores indicate stronger agreement.

^bThe asymptotic 95% CI (confidence intervals) found significant agreements between patient and caregiver ratings on 16 out of 17 items.

^cThe one non-significant (NS) CI between patient and caregiver ratings was the IPOS practical issues item (k = .17).

On the IPOS open-ended question, there were additional palliative care problems reported by patients and caregivers. These responses revealed each family’s specific palliative care issues such as patients being unable to do things they had been able to do before (walk, exercise), and financial problems. Additional patient problems reported by caregivers were patient edema, poor sleep patterns, behavioral issues, poor speech/communication, agitation, and anger.

Results of Question 3.

Does the patient IPOS total score predict the patient HF health status and the family caregiving burden scores?

Means (M) and standard deviation (SD) of patient and caregiver dependent (outcome) and independent (predictive) variables are described in Table 4. The patient IPOS total score was entered into the regression analysis as the predictive variable for patient HF health status. The family caregiving burden scores were entered as the dependent (outcome) variables (Table 4). The regression analysis of the patients’ IPOS total score predicting patients’ HF health status (KCCQ) scores was statistically significant (R² = .25, F = 5.93, P < .05; β = −.50, P < .05). Also, the regression analysis of patients’ IPOS total scores predicting caregiving burden (ZBI-12) scores was statistically significant (R² = .32, F = 8.59, P < .01; β = .57, P < .01). A positive beta weight (β) indicated a positive relationship between the IPOS and the caregiving burden scores. A negative beta weight (β) indicated a negative relationship between the IPOS and the KCCQ scores.

Table 4.

Means, SD, and Range of Patient IPOS Total Score, Patient HF Health Status and Caregiving Burden Scores.

Variables	M (SD)	Range of Scores
Dependent (outcome) variables
Patient HF health status (KCCQ summary score)	46.02 (28.57)	0 to 100
Caregiving burden score (ZBI-12)	17.79 (9.94)	0 to 48
Independent (predictive) variable
Patient IPOS total score	22.86 (10.83)	0 to 68

Note. There was no association between demographic variables with these dependent (outcome) measures. The independent (predictive) variable of patient IPOS total score did predict both dependent (outcome) variables.

Discussion

This study addressed the gap in information related to measuring palliative care needs for patients with HF and vascular dementia. The agreement between these patients and their caregivers IPOS item rating scores were consistent in identifying patients’ physical, emotional, and palliative care communication and practical concerns for dementia and HF conditions. Specific problems rated by both patients and caregivers on the IPOS were patients’ physical symptoms (e.g., weakness, shortness of breath, poor mobility, poor appetite, and drowsiness); emotional problems (patient and caregiver anxiety, depression, and feeling at peace); and communication/practical issues (sharing feelings, information, and practical palliative care concerns).¹⁹

Most of these findings suggest that the patient experiences are congruent with their caregiver observations except for patients’ pain and information needs. It is common for caregivers to report greater information needs than patients. Many caregivers perceive the patient’s pain to be more concerning than the patient. Providing simple pain assessment tools such as facial pain scales and behavioral non-verbal observational pain scales help caregivers to quantify the patient’s pain more accurately.⁴⁴ Importantly, if there are IPOS items that do vary between patients and caregivers, this alerts healthcare professionals to assess any causes of the differences.

Being able to predict patient HF health status and caregiving burden allows for further assessment of palliative care needs. The extensive worldwide use and testing of the KCCQ support interpreting scores below the midpoint as indicating poor HF health status.³⁶ Notably, the regression results found that the patients’ IPOS total scores predict both patients’ HF health status and caregivers’ burden measures. Thus, IPOS results can contribute to multidisciplinary medical decisions.

In this sample, both patient and caregiver responses agreed on IPOS physical, emotional, and communication/practical concerns. Clinicians use the IPOS measures to identify palliative care needs across time, thus managing changing care needs.³⁷ The IPOS measure is often used repeatedly¹⁸ to follow patient HF health status changes and specific caregiver burden needs. The open-ended question on the IPOS allowed identification of additional symptoms or concerns specific to each family. Several caregivers reported need for support to ease their home care burden (e.g., respite care, time for themselves).

HF and vascular dementia are devastating for patients and their family caregivers.⁶ Thus, measuring patients’ symptoms and palliative care needs using the IPOS can guide palliative home management and lead to caregivers’ preparedness.²⁸ Yet, across several studies, caregivers reported they were not prepared for patients’ problems associated with dementia, the patient’s illness condition, and preparing for end-of-life challenges.^9,11,12 Palliative care helps patients with physical care, comfort, symptom control, and preparing for home care.^28,45

Caregivers can prepare for home dementia care challenges by maintaining open communication with healthcare providers, building a support network, taking care of their own health, and utilizing available resources.^24,28 Those resources may include financial, social support referrals, and sharing caregiving responsibilities among other family members.⁴⁶ Nurse-led interventions for in-home palliative care for HF are highly regarded.^47,48

Clinicians and researchers could use the IPOS to monitor changes across time. The changing item ratings on the IPOS can alert multidisciplinary healthcare providers to increased severity of symptoms, poorer communication, and worsening practical care concerns. Repeated clear communication between caregivers and healthcare providers also improves preparedness. Preparedness information, communication improvements, and supporting psychological adjustments prepares caregivers for meeting patient palliative care needs when patients health is declining.⁴⁹

Recently, the IPOS for dementia was verified and became available. This version includes additional dementia-related items, for examples, swallowing problems (e.g., chokes, holds food in mouth), difficulty communicating, sleeping problems, skin breakdown, hallucinations and/or delusions.⁵⁰ Some of these same problems and concerns were reported in this study’s open-ended questions of the IPOS by patients and their family caregivers. Also, family caregivers and healthcare professionals have completed the IPOS for the assessment of patients with cognitive decline in dementia management clinics.^16,51 Further, in this study, the IPOS has successfully assessed the palliative care needs of HF patients with cognitive declines.

Limitation

This study has limited generalizability only to family caregivers providing home palliative care for patients with HF and dementia. This descriptive design has small sample size limitations. Another limitation of the study was the sample lacked diversity of ethnicity.⁵² This sample was consistent with the West Virginia state-wide census, where over majority of the population (>93%) is non-Hispanic White.⁵³ Palliative care support for non-white caregivers of people living with dementia is recommended to improve health equity.⁵⁴ However, the unique benefit of this study was the paired data collection, and the significant agreement between patient and caregiver IPOS ratings were established.

Notably, the low alpha reliability score (.64) was found for communication/practical subscale. There are several possible causes of alpha below .70. Clinical decision based on these communication/practical items may be limited. In addition, the low alpha may indicate that the three items under communication/practical issues subscale are measuring different constructs. The alpha reliability score was improved from .64 to .70 after removing the practical problems item. Further investigation in a large sample is warranted.

Conclusion

Our results support other researchers’ findings that IPOS could be accurately completed by these patients with HF and vascular dementia. The IPOS did identify congruence between patient and caregiver ratings of patient’s physical, emotional symptoms and palliative care needs. Using IPOS scores, patients’ HF health status and caregiving burden were also identified. Further, patient and family caregiver responses to the open-ended questionnaire can provide insight into specific areas of their palliative care needs. Managing palliative home care based on clinically relevant scientifically verified IPOS measures is recommended.

Abbreviations

HF

Heart failure

IRB

Institutional review board

IPOS

integrated palliative care outcome scale

KCCQ

Kansas city cardiomyopathy questionnaire

ZBI

Zarit burden inventory (caregiving burden scale)