Caregivers’ Heart Failure Knowledge is Necessary but not Sufficient to Assure Engagement with Patients in Self-care Maintenance

Background

I was making a routine hospice home visit to one of my favorite couples, Joe and Shirley. We sat at their kitchen table and chatted about their week as I filled Joe’s medication box. Joe had advanced heart failure and so took between 14–21 different pills a day depending on his weight that day. He had been successfully filling his own medication boxes for about 10 years but was starting to make mistakes. Shirley didn’t seem willing to take on this responsibility. I couldn’t understand her hesitance at first. However, as I got to know them better I discovered that when Joe and Shirley talked about handling Joe’s heart failure care at home they always argued. This caused quite a bit of distress for this long time couple. So Shirley just stayed out of it. But this refusal to fill her husband’s medication box had negative repercussions. Joe had recently experienced multiple hospitalizations due to symptom exacerbation related to his own medication errors. More importantly this lead to a serious decline in his condition and quality of life. It was possible that Joe could live for several more years with careful heart failure self-care at home but he needed help.

Palliative care has been inherently patient and family centered since inception.¹ Yet relatively little is known about the impact that families have on patient’s heart failure (HF) self-care at home especially when compared with oncology and dementia populations.² Self-care is defined as a naturalistic decision-making process that results in daily HF care such as adherence to recommendations for diet, medications, and exercise and weight monitoring, as well as symptom management if and when symptoms occur.³ For the purposes of this paper, self-care refers to both the patient’s care for his/her HF OR the caregiver’s care for the patient’s HF.

HF is a growing population in palliative care.¹ Hospice and palliative nurses are taking care of more HF patients and families and need to be familiar with the HF self-care evidence base. What is currently known is that HF patients are rarely proficient in self-care. ^3,4 One suggested solution has been to engage the help of an informal caregiver. For this reason, HF patient/informal caregiver dyads are currently undergoing careful examination in the cardiovascular nursing literature.^5–8 Recent dyadic studies have uncovered gradients of individual contributions and expertise in dyadic self-care, ⁵ incongruence between patient and caregiver perceptions of self-care ^6,8 and the impact of self-care on dyadic quality of life.⁷ This new focus on HF dyads comes with an associated need to study how dyadic dynamics influence self-care in greater depth. With the current emphasis on increasing access to outpatient palliative care it is imperative that palliative care nurses begin to assess and intervene, if necessary, in dyadic HF care.

Our team developed the Dyadic Heart Failure Care Typology ⁹ to help clinicians understand and assess dyads. Four care types: patient, caregiver, collaborative, and complementary suggested by the Actor Partner Interdependence Model,¹⁰ interdependence theory,¹¹ and clinical practice were presented and characterized in an earlier qualitative study.⁹ Using instrumental case study methodology, defining features, subtypes, characteristics, and context were developed for the four care types. In particular, two of the care types (patient and caregiver types, where patients or caregivers provide care without input from the other partner) were determined to be individually-oriented while the other two types (collaborative and complementary where patients and caregivers provide care together) are relationally-oriented. An instrument to measure the care types, the Dyadic Symptom Management Type (DSMT) scale, was also developed and tested as part of this process.¹² In preliminary psychometric evaluation, the DSMT scale showed fair strength of agreement between patient and caregiver response when both partners were tested and dyads expressed no difficulties in understanding the item or selecting their particular type.¹² The next logical step after this rigorous process was to examine the relationship between dyadic orientation and patient reported outcomes.

The chronic illness literature suggests that dyads who share decisions and activities, such as relationally-oriented types, are more efficient and effective than those who don’t.^13,14 Therefore, the purpose of this study was to examine differences between individually- and relationally-oriented dyads relative to patient symptom management scores, controlling for HF knowledge, relationship quality and the nature of the patient/caregiver relationship (spousal vs. nonspousal).

Methods

Study design

This study was part of a cross-sectional prospective study investigating dyadic symptom management in HF (R03NR014524). The conceptual framework for this study is the Actor Partner Interdependence model (APIM).¹⁰ Dyadic data, whether spousal, adult child, or other relationship are highly correlated. Therefore, dyadic data violates the assumption of independence necessary for statistical analysis. The APIM accounts for that lack of independence by modeling the interdependence of both the independent variables and the unexplained variance in the dependent variables from both members of the dyad resulting in more accurate estimates of parameters.¹⁰ In this project, dyadic type (measured in both patient and caregiver) was the independent variable and self-care (measured in both patient and caregiver was the dependent variable).

Setting and Sample

Patient/caregiver dyads were recruited from a large academic medical center in the northeast United States. After obtaining institutional review board approval from Penn State Hershey Medical Center (IRB# STUDY00000015), patients and caregivers were recruited by trained research assistants (RAs). Caregivers were identified by the patient as someone who provided care without payment. To be included in the study, both partners had to be 21 or older, willing to participate, and the patient had to have a documented HF diagnosis in the electronic health record. Post heart transplant or left ventricular assist device patients were excluded to assure a homogeneous self-care experience.¹⁵

Procedure

Daily hospital and outpatient clinic records were screened. If patients met the inclusion criteria, the dyad was approached by the RA, who explained the study and obtained written consent before administering the instruments via pencil and paper.

Measures

Personal and dyadic characteristics

Patient and caregiver self-reported age, type of dyadic relationship (spousal, adult child, or other), gender, race/ethnicity, education level, and income adequacy with a study specific questionnaire. Patient HF diagnosis was confirmed in the electronic health record.

Dyadic Symptom Management Type

The Dyadic Symptom Management Type (DSMT) scale contains one criterion-referenced item derived from the dyadic HF care typology.¹² The scale gives a definition of self-care at a 9^th grade readability level and then directs the selection of one and only one statement that best describes how the individual and their care partner take care of the patient’s HF. For example, for the patient dyadic type, the patient selects the statement “I take care of the majority of my heart failure without any input from my caregiver.” If the dyad is congruent (agrees on dyadic type) the caregiver selects the statement “My care recipient takes care of the majority of his/her heart failure without any input from me.” The caregiver dyad type statement is similarly worded for the patient: “My care partner takes care of the majority of my heart failure without any input from me.” In the two relationally-oriented dyadic types the item uses we rather than I or his/her statements. Evidence for preliminary validity have been reported previously.^9,12 To answer our question we collapsed the patient and caregiver types into a category titled individually-oriented and collapsed the collaborative and complementary types into a category titled relationally-oriented. We then adjudicated a dyad to be congruent if both patient and caregiver chose the same dyadic type orientation, either individual or relational. For example, if the patient selected collaborative and caregiver selected complementary they were adjudicated as relationally congruent. Both the patient and caregiver completed the measure.

Dyadic Adjustment Scale-short form (DAS-7)

The DAS-7 is a 7 item self-report measure of relationship quality that assesses dyadic consensus, cohesion, and global satisfaction.¹⁶ For example, respondents are asked to indicate the degree of agreement (always, almost always) or disagreement (occasionally, frequently, almost always, and always) that they and their care recipient experience in philosophy of life, aims, goals, and amount of time spent together. Relationship quality is a well-known mediator and is frequently used as a covariate in couples’ research¹⁷. The DAS-7 exhibits good internal consistency (average α =.80) and discrimination¹⁶ and has been used in multiple studies in clinical settings.^18–20 Both the patient and caregiver completed the measure and an average score across both members was used in the analysis.

Dutch Heart Failure Knowledge Scale (DHFK)

The DHFK is a 15 item measure of general HF knowledge, recognition of HF symptoms, and HF treatment knowledge. The DHFK is a multiple choice test with a possible range of 0 (no knowledge) to 15 (optimal knowledge). As an example, one of the DHFK questions shows a picture of a soup can label and then asks the test taker to select the number of servings in the can and how much sodium is in one serving. The DHFK has a published α=.62 which is not unexpected in a multiple choice test and is able to discriminate between uneducated and educated HF patients.²¹ Both the patient and caregiver completed the measure and individual scores for each member were used in the analysis.

Self-care of HF Index (SCHFI)

The SCHFI v.6 is a gold standard instrument measuring heart failure self-care. It includes three scales – the Self-care maintenance scale (10 items), Self-care management (6 items), and Self-care confidence or self-efficacy (6 items). For example, a self-care maintenance item asks the patient how often (never or rarely, sometimes, frequently, or always or daily) they weigh themselves; whereas a self-care management item asks them how quickly they recognized increase ankle swelling as a symptom of HF. A higher score indicates better self-care. Confirmatory factor analysis supports construct validity. ²² Coefficient alphas have been reported to range from .55, .70 and .83 for self-care maintenance, self-care management, and self-care self-efficacy respectively.²³ Patients were measured using the SCHFI. Caregivers were measured by an adaptation of the patient scale used in multiple other studies and reporting similar psychometric properties.^7,24–26 This instrument mirrors the patient SCHFI but asks the caregiver to indicate how routinely they observe or help their care recipient perform self-care maintenance or self-care management, and how confident they are that they can help their care recipient take care of their HF. Individual scores for each were used.

Data Analysis

Personal and dyadic characteristics were summarized using descriptive analysis. A paired t-test was used to compare dyad members on relationship quality and HF knowledge.

The Actor-Partner Interdependence Model (APIM) with distinguishable dyads was used to determine the effect of dyadic symptom management type on patient symptom management. The APIM model allows for the estimation of regression parameters within the nested design (i.e., dyad members nested within dyad pair) and allows the inclusion of both individual- and dyad-level predictors. Covariates in the model included HF knowledge, relationship quality and type of dyad relationship (dichotomized as spousal vs. non-spousal because of unbalanced groups). For relationship quality, there was a high degree of consistency between the patient and caregiver ratings and so an average (across both dyad members) DAS-7 score was used in the analysis. HF knowledge, however, was significantly different across dyad members and individual scores were used. All analyses were conducted in SAS 9.4; an alpha level of .05 was used throughout.

Results

Sample Description

Fifty five dyads (n=110 individuals) contributed data. On average, patients were older, financially comfortable, white males (Table 1). Caregivers were younger and female. Most dyads were spousal (63.6%; n = 35), adult child (29.1%; n = 16) or other (7.3%; n = 4). Thirty seven dyads selected relationally- oriented types, 14 dyads selected disparate types and 4 dyads selected individually- oriented types. Descriptive statistics on the instruments are found in Table 2.

Table 1.

Sample Demographic Characteristics

Characteristics	Patient (n=55)	Caregiver (n=55)
	Mean (SD) or proportion	Mean (SD) or proportion
Age (in years)	74.8 (10.09)	62.7(14.9)
Gender (% Female)	33%	78%
Race/Ethnicity
White	54 (98.1%)	54 (98.1%)
Black	1 (1.9%)	1 (1.9%)
Education
HS or less	35 (63.6%)	29 (52.7%)
Some college	7 (12.7%)	11 (20%)
College or advanced degree	13 (23.6%)	15 (27.3%)
Income to make ends meet
More than enough	12 (22.2%)	14 (25.9%)
Enough	32 (59.3%)	31 (57.4%)
Not enough	10 (18.5%)	9 (16.7%)

Note.

^†One dyad (one patient and one caregiver) left this item missing

Table 2.

Patient and Caregiver Self-Report Descriptive Statistics

	Patient			Caregiver			Total
	f (%)			f (%)			f (%)
Dyadic Symptom Management Type
Patient	8 (14.6%)			9(16.4%)			17 (15.4%)
Caregiver	2 (3.6%)			3 (21.8%)			5 (4.5%)
Collaborative	27 (49.1%)			26 (47.3%)			53 (48.1%)
Complementary	18 (32.7%)			17 (30.9%)			35 (31.8%)
Incongruent dyads							14 (25.5%)
							Test for differences
	Mean	N	SD	Mean	N	SD	t	p-value
DAS-7	24.89	55	4.86	25.2	55	4.57	−0.54	.589
DHFK	12.18	55	2.29	13.29	55	1.94	−3.25	.002
Self-care Maintenance	66.92	55	13.75	54.47	55	15.46	4.99	<.001
Self-care Management	56.10	41	25.12	61.08	41	20.90	−1.11	.275
Self-care Self-efficacy	55.28	55	18.52	58.06	55	21.85	−0.90	.371

Note.

^†As frequently reported in other studies, fewer dyads completed the Management scale due to the requirement that dyads recognize symptoms in the previous month.

Preliminary analysis demonstrated that whether the dyads were individually- or relationally-oriented did not make a difference in patient self-care management scores. None of the measured variables were significantly related to either the patient’s or the caregiver’s self-care management scores. What was found, instead, was that caregivers in incongruent dyads (i.e. dyads who disagree on dyadic type) who scored lower on HF knowledge had significantly lower maintenance scores compared to those in congruent dyads with higher HF knowledge. Finally, relationship quality and dyadic type were significantly related to dyads’ scores on self-care self-efficacy. Better relationship quality in caregivers was related to higher scores on self-care self-efficacy. Patients in spousal dyads had higher scores on self-care self-efficacy compared to patients in non-spousal dyads. The final APIM models appear in Table 3.

Table 3.

APIM results

	Maintenance		Management		Confidence in care
	Patient	Caregiver	Patient	Caregiver	Patient	Caregiver
Intercept	66.909 (2.092)	67.418 (2.114)	54.665 (4.234)	55.747 (4.335)	52.029 (3.01)	55.389 (3.059)
Congruent (ref = Congruent)	−3.9 (4.144)	−8.469 (4.142)	0.955 (8.383)	−1.166 (8.783)	4.923 (5.97)	9.611 (5.965)
Relationship quality	−0.036 (0.448)	0.805 (0.452)	−0.433 (0.899)	1.407 (0.911)	−0.022 (0.645)	1.696 (0.653)
Type of dyadic relationship (ref = Spousal)	3.165 (1.986)	1.462 (1.947)	4.521 (4.116)	0.614 (4.131)	6.366 (2.892)	−1.425 (2.806)
HF knowledge	0.976 (0.627)	1.486 (0.725)	−1.258 (1.461)	−1.005 (1.775)	1.893 (1.114)	1.504 (1.289)

Note. Bold estimates p < .05.

Conclusions

This study examined the relationship between individually- vs. relationally-oriented dyadic symptom management types and patient symptom management. We found that the type of dyad was not as important as that the dyad agreed on their particular type. Incongruence (i.e. patient and caregiver disagree on who is providing self-care) was found to be a more important factor in whether caregivers engage in adherence to the plan of care and monitoring the patient (i.e. self-care maintenance) than whether they worked alone or together with the patient on these tasks. Impeccable self-care maintenance is necessary to prevent HF exacerbations and subsequent hospital admissions.²⁷ Therefore nurses, in general, and palliative care nurses, in particular, need to understand the facilitators and barriers to HF caregiver engagement and then leverage that knowledge to improve engagement and the caregiving experience overall.

Dyadic incongruence has been identified by multiple investigators as a new construct with major impact on outcomes.^6,8,28 Dyadic incongruence has previously been examined qualitatively ^6,8 but our work is the first, to our knowledge, to measure HF self-care incongruence quantitatively. In doing so, we found the approximately 25% or 1 in 4 dyads were incongruent. If this new finding is supported in future studies it has profound implications. It is possible that when a patient indicates that his caregiver provides the majority of his care on the DSMT but his caregiver indicates that the patient provides the majority of his care, that neither are actually providing any care. This finding needs to be examined further, but as it stands, it gives one pause. In an example from our study of the potential pitfalls of dyadic incongruence, a patient (mother) indicated that the caregiver (daughter) provided all of her HF self-care but the daughter indicated that they collaborated on the care. Both members of the dyad showed a low level of HF knowledge despite a 10 year history of HF. The caregiver’s score on the maintenance scale was far below the 70 which is generally accepted as adequate. When first asked what she did to care for her mother’s HF on a daily basis she reported that she merely watched her mother’s diet but did not manage her medications or contribute to other maintenance activities such as daily weights or encouraging healthy living. When the daughter was then asked how she responded to any changes in her mother’s condition, she reported that instead of taking any action when she noticed that her mother was becoming symptomatic (i.e. dependent edema) she just called the doctor. This lack of initiative on the part of the caregiver to either prevent or manage symptoms has profound implications for nurses particularly as chronic illness management, such as HF management, is increasingly falling under the purview of nursing. Earlier studies linked incongruence to frustration, distress ⁶ and poorer self-care outcomes.⁸ What our study adds is that incongruence results in 1 in 4 caregivers reporting measurably less engagement in day to day HF care such as maintaining a 2 gram sodium diet, encouraging the patient to take cardiac medications and perform daily weights. Further examination in larger, more diverse samples is needed to understand the full impact of dyadic incongruence on patient outcomes. However, this preliminary data suggests that addressing dyadic incongruence, a potentially modifiable situation, may offer new ways to improve patient’s HF self-care while also improving caregivers’ experiences of stress and burden.

As in an earlier study,²⁵ relationship quality was significantly related to dyads’ self-care. In that earlier study, however, relationship quality was measured by a 1-item question asking the person to judge how good their relationship was. This one item was linked to self-care maintenance and self-care management.²⁵ What our study suggests is that that self-care self-efficacy may be moderating the relationship between relationship quality and HF self-care, at least for the caregiver. The impact of the caregiver’s perception of the relationship on self-efficacy is supported in the literature.²⁹ What we found, using a gold-standard relationship quality scale, the DAS-7, was that the better the caregiver’s perception of the quality of the relationship, the higher their scores on self-care self-efficacy. It is possible, for example, that when an adult child caregiver and their parent have a better relationship the caregiver is able to devote time and energy to gaining mastery in self-care rather than in dealing with the interpersonal conflict. This improved mastery then results in improved self-care self-efficacy which in turn results in better self-care. This association between relationship quality and self-care was not found in patients. However, patients in spousal dyads had higher scores on self-care self-efficacy compared to patients in non-spousal dyads suggesting some relationship dynamic that was not captured by the DAS-7. Further examination is needed in this area.

Limitations

Our sample is primarily white, spousal, economically comfortable, relatively small, and accrued from a single academic medical center. It is possible that some relationships would have been significant given a larger sample or that these relationships will change with different covariates. Longitudinal studies are also needed to understand the intra-individual variations in dyadic self-care over time.

Implications

This study has important specific implications for palliative nursing practice. A current assumption in palliative care is that informal caregivers will automatically engage in day to day care. Our study adds a significant stipulation to this assumption – caregivers are more likely to engage in day to day HF care if they agree with the patient on who is providing that care and are fairly knowledgeable about HF. If they disagree with the patient and are less knowledgeable about HF then they are less likely to engage in this necessary day to day care. This becomes a critical factor as the patient’s condition begins to deteriorate and he/she is unable to provide that care for themselves at home. If the caregiver does not step in, who will? Will the patient be able to safely remain at home if that is what they desire? These are questions which hospice and palliative care nurses deal with on a daily basis in their clinical practice.

The DSMT scale is a one item question that can be administered to both patient and caregiver in under 2 minutes in the hospital, clinic, or home. The DSMT scale can be used in two very important ways. First, using the DSMT scale, a busy nurse can identify which person is providing the HF care at home and then target any education to the person actually providing the care resulting in improved uptake and implementation of the information. Secondly, the DSMT scale can also be used as a quick assessment tool for the presence of incongruence. By comparing both scales, a nurse can identify incongruent dyads and develop a plan of care that includes putting in place additional structures and processes of care to help these patients and caregivers achieve their long term goals.