Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2020 Sep 1.
Published in final edited form as: Aging Ment Health. 2018 Dec 20;23(9):1122–1129. doi: 10.1080/13607863.2018.1481923

Relationship and Communication Characteristics Associated with Agreement between Heart Failure Patients and Their CarePartners on Patient Depressive Symptoms

Erin D Bouldin 1,*, James E Aikens 2, John D Piette 3,4, Ranak B Trivedi 5,6
PMCID: PMC6586543  NIHMSID: NIHMS1512329  PMID: 30569750

Abstract

Objectives:

Informal caregivers who recognize patients’ depressive symptoms can better support self-care and encourage patients to seek treatment. We examined patient-caregiver agreement among patients with heart failure (HF). Our objectives were to (1) identify distinct groups of HF patients and their out-of-home informal caregivers (CarePartners) based on their relationship and communication characteristics, and (2) compare how these groups agree on the patients’ depressive symptoms.

Method:

We used baseline data from a comparative effectiveness trial of a self-care support program for veterans with HF treated in outpatient clinics from 2009–2012. We used a cross-sectional design and latent class analysis (LCA) approach to identify distinct groups of patient-CarePartner dyads (n=201) based on relationship and communication characteristics then evaluated agreement on patients’ depressive symptoms within these groups.

Results:

The LCA analysis identified four groups: Collaborative (n=102 dyads, 51%), Avoidant (n=33 dyads, 16%), Distant (n=35 dyads, 17%), and Antagonistic (n=31 dyads, 15%). Dyadic agreement on the patients’ depressive symptoms was highest in the Distant (Kappa (κ)=0.44, r=0.39) and Collaborative groups (κ=0.19, r=0.32), and relatively poor in the Avoidant (κ= –0.20, r=0.17) and Antagonistic (κ=–0.01, r=0.004) groups. Patients in Avoidant (61%) and Antagonistic groups (74%) more frequently had depression based on self-report than patients in Collaborative (46%) and Distant (34%) groups.

Conclusion:

Caregiver relationships in HF tend to be either Collaborative, Avoidant, Distant, or Antagonistic. Patients’ depressive symptoms may negatively affect how they communicate with their caregivers. At the same time, improved patient-caregiver communication could enhance dyadic consensus about the patient’s depressive symptoms.

Keywords: caregivers, heart failure, depression, communication, proxy

Background

An estimated 5.7 million adults in the US have chronic heart failure (HF) (Mozaffarian et al., 2016). Heart failure may be diagnosed at any age but typically occurs among older adults (Mozaffarian et al., 2016). It is a progressive disease that is associated with lower quality of life and earlier mortality, though survival has increased substantially over the past 40 years (Levy et al., 2002; Mozaffarian et al., 2016). HF-related hospitalizations are common and contribute to the high costs of the disease (Agarwal et al., 2016; Blecker, Paul, Taksler, Ogedegbe, & Katz, 2013; Mozaffarian et al., 2016). Self-care for HF involves taking medications as prescribed, limiting sodium and fluid intake, and monitoring symptoms like edema, weight gain, and shortness of breath (Artinian, Magnan, Sloan, & Lange, 2002; Moser & Watkins, 2008). Optimal self-care is essential for HF management, and can reduce exacerbations and hospitalizations, improve longevity, and increase quality of life (McAlister, Stewart, Ferrua, & McMurray, 2004; Moser & Watkins, 2008). However, HF self-care is significantly complicated by depression, which is experienced by 1 in 5 HF patients (Rutledge, Reis, Linke, Greenberg, & Mills, 2006). Depression in HF patients is associated with higher risks of hospitalization and death (Jani et al., 2016; Rutledge et al., 2006), in part because depressive symptoms reduce self-care (Buck, Mogle, Riegel, McMillan, & Bakitas, 2015; Janevic, Rosland, Wiitala, Connell, & Piette, 2012; Moser & Watkins, 2008; Riegel et al., 2009). Although the mechanisms are still being explored, depression may interfere with self-care by reducing self-care confidence (Chang, Wu, Chiang, & Tsai, 2017), particularly among HF patients lacking social support (Graven et al., 2015; Lee, Lennie, Yoon, Wu, & Moser, 2017). Therefore, recognizing and managing depression is an important aspect of maintaining health and quality of life among patients with HF, and informal caregivers may be uniquely positioned to do so.

Informal caregivers – family and friends who help in the community – provide a broad range of supports to patients with chronic conditions, ranging from helping manage medications, to assisting with cooking, and aiding with activities of daily living like dressing and bathing (Trivedi et al., 2016; Joo, Fang, Losby, & Wang, 2015; Talley & Crews, 2007). The involvement informal caregivers in self-care has positive effects on patient self-care and clinical outcomes, in part because they monitor the health and well-being of patients (Sayers, Riegel, Pawlowski, Coyne, & Samaha, 2008; Strömberg, 2013). Informal caregivers who recognize patients’ depressive symptoms may be able to better support their self-care activities and encourage patients to seek appropriate depression treatment (Janevic et al., 2012).

The extent to which caregivers can effectively monitor their patients depends in part on the quality of their relationship and the effectiveness with which patients and caregivers communicate with each other (Mayberry & Osborn, 2014; Retrum, Nowels, & Bekelman, 2013; Rosland, Heisler, Choi, Silveira, & Piette, 2010; Rosland, Heisler, & Piette, 2012). However, many caregivers report barriers to discussing the patient’s health (Janevic et al., 2012; Retrum et al., 2013), or have negative interactions characterized by criticism, over-involvement, or avoidance (Manne, 1999; Rosland et al., 2010; Vitaliano, Young, Russo, Romano, & Magana-Amato, 1993). Perhaps for these reasons, correlations between patient and proxy reports of psychological health are often found to be weak (Neumann, Araki, & Gutterman, 2000). In one study, informal caregivers attributed more depressive symptoms to HF patients than those reported by patients themselves (Quinn, Dunbar, & Higgins, 2010). Few studies have directly evaluated informal caregivers’ accuracy in recognizing depression, or identified characteristics associated with accurately reporting patients’ depressive symptoms. In particular, the ability of out-of-home caregivers to recognize depression in patients remains largely unexamined.

In this study, we focused on CarePartners, defined as family members or friends who do not live with the HF patient and who are willing to become involved with the patient’s self-care (Piette et al., 2008; Piette, Striplin, Marinec, Chen, Trivedi, et al., 2015; Piette, Striplin, Marinec, Chen, & Aikens, 2015). We were interested in evaluating how closely CarePartners’ reports of patient depressive symptoms aligned with patients’ own reports and in identifying factors associated with higher agreement between patients and CarePartners, focusing on relationship and communication characteristics based on the existing literature. Therefore, we characterized patient-CarePartner dyads based on their relationship and communication characteristics using a latent class analysis approach. We also determined whether relationship and communication are associated with the agreement between patients and CarePartners on patients’ depressive symptoms. The results of this study could inform programs for HF patients and their CarePartners to improve the recognition of depressive symptoms and ultimately depression and HF self-management among patients.

Methods

Sample

We used data from a comparative effectiveness trial of male veterans with HF treated in the Veterans Health Administration (VHA) Cleveland Medical Center outpatient clinics from 2009–2012 (Piette, Striplin, Marinec, Chen, Trivedi, et al., 2015; Piette, Striplin, Marinec, Chen, & Aikens, 2015). The 12-month trial originally included 369 HF patients who had a VA primary care provider, had a VA outpatient visit in the past year, could complete phone calls in English, and identified a CarePartner (Piette et al., 2008; Piette, Striplin, Marinec, Chen, Trivedi, et al., 2015; Piette, Striplin, Marinec, Chen, & Aikens, 2015). Patients who lived in skilled nursing or received palliative care, were prescribed supplemental oxygen, or had been diagnosed with dementia, bipolar disorder, or schizophrenia were not eligible for the study. At baseline, patients and CarePartners rated aspects of their relationship and communication with one another and their own health. Additionally, CarePartners rated the patient’s health including depressive symptoms. The present study included the 201 dyads in which both the patient and CarePartner rated the patient’s depressive symptoms at baseline, and excluded 168 dyads in which the CarePartner did not rate the patient’s depressive symptoms.

Measures

Depressive Symptoms

We measured depressive symptoms using the 10-item version of the Centers for Epidemiologic Studies Depression scale (CES-D-10) (Carpenter et al., 1998; Irwin, Artin, & Oxman, 1999; Kohout, Berkman, Evans, & Cornoni-Huntley, 1993). Items ask respondents to report how frequently during the past 4 weeks they felt depressed, happy, lonely, sad, and disliked, and also how often they experienced restless sleep, enjoyed life, could not get going, felt everything was an effort, and felt people were unfriendly. Response options ranged from “rarely or none of the time” (0) to “most or all of the time” (3). This measure was completed by patients for their own symptoms, and the wording was adapted so that CarePartners could rate the patient’s symptoms.

We examined both the continuous summed CES-D-10 score and a dichotomous measure of depressive symptoms defined as a score was ≥10, which is similar to using a cutpoint of 16 on the full 20-item CES-D (Ƙ=0.97) (Andresen, Malmgren, Carter, & Patrick, 1994). The CES-D-10 has moderate to good test-retest reliability in the short term (r=0.71) and over one year (r=0.59) and good validity based on correlations with other measures (Andresen et al., 1994).

Dyad Agreement on Patient Depressive Symptoms

We considered patients and CarePartners to be in agreement when scores from both members of the dyad resulted in the same classification of the patient’s depression symptoms (i.e., both scores classified patient as depressed [CES-D-10≥10] or both scores classified patient as not depressed [CES-D-10<10]). As a sensitivity analysis, we considered whether the continuous CES-D-10 scores were similar within dyads.

Relationship characteristics

We measured relationship characteristics using 17 items that assessed how close the patient felt to the CarePartner, the frequency with which patients experienced negative emotions when talking with their CarePartners, and the amount of caregiving-related strain CarePartners experienced. Patients were asked to rate how close they were to the CarePartner using a scale from 1 (not very close at all) to 10 (very close). Patients were asked to report how much they experienced the following six emotions when they talked with their CarePartner during the past 3 months: sadness, loneliness, anger, tension, guilt, and frustration. Patients scored each item on a scale of 0 (none) to 3 (a great deal). Finally, we measured CarePartner strain using the modified 10-item Caregiver Strain Index (Thornton & Travis, 2003). CarePartners were asked to report whether helping the patient affected them in a variety of areas, including sleep, physical strain, and emotional tensions. For each type of strain, CarePartners could report, “no”; “yes, sometimes”; and “yes, on a regular basis.”

Communication

We measured communication between patients and CarePartners using 11 items that assessed each persons’ feelings about their contact frequency and how their conversations about HF typically progressed. Patients reported on their agreement with the following HF-related communication statements: (1) “It is important for you to talk with your CarePartner about your illness,” (2) “You would like to talk with your CarePartner about your illness more than you have done,” and (3) “It is difficult for you to talk to your CarePartner about your illness.” Patients rated the statements on a 5-point scale from “strongly disagree” to “strongly agree.” Patients also reported on how they typically experienced specific feelings when talking about their HF: “When you and your CarePartner talk about your health, how often would you say the following happens?” (1) “I don’t want to be asked too many questions because it’s none of their business,” (2) “I’m frustrated because they always ask the same questions,” and (3) “It’s depressing for me.” Similarly, we asked CarePartners, “When you and your [relative/friend] talk about their health, how often would you say the following happens?” (1) “I don’t want to ask too many questions because it’s none of my business,” (2) “I’m frustrated because they always bring up the same problems,” and (3) “It’s depressing.” Response options included never, rarely, sometimes, mostly and always. We included two measures of communication frequency, both reported by the patient. Patients reported how many times they talked to the CarePartner in an average week over the past 6 months and how many times they saw their CarePartner in person in an average month over the past 6 months.

Categorical Variables for Modeling

We considered patients to feel close to their CarePartner if they reported a score of 8 or higher (out of a maximum 10) based on the distribution of scores in the sample (mean=8.3; median=9). We classified patients as experiencing negative emotions if they reported experiencing at least one of the six emotions regularly, which we defined as scoring 2 or 3 (on a scale from 0 [none] to 3 [a great deal]). We classified CarePartners as experiencing strain if they said, “yes, on a regular basis” in at least one of the ten areas. For each item rated on a scale from strongly disagreed to strongly agreed, we dichotomized responses into patients who somewhat or strongly agreed with a statement versus those who strongly disagreed, somewhat disagreed or were neutral. For each item rated on a scale from never to always, we created a dichotomous variable to indicate patients mostly or always reporting an experience versus sometimes, rarely, or never experiencing it. We created variables to indicate whether patients said they talked to the CarePartner two or more times per week in an average week over the past 6 months and whether they saw their CarePartner at least once per week in an average month over the past 6 months.

Covariates

Patients and CarePartners reported their age, gender, race/ethnicity category, highest level of education, and whether they were employed. Patients reported how they were related to the CarePartner (e.g., friend) and how far they lived from the CarePartner. CarePartners were asked to think about all the kinds of help they provide the patient (specifically, shopping and errands, household chores or preparing meals, taking prescription medications, transportation, managing finances, and arranging medical services) and to report how many hours per week, on average, they spent providing help to the patient. In a separate question, CarePartners were asked to report how many hours they spent in a typical week providing support and reassurance to the patient by phone.

Analysis

Because we were interested in grouping patients and their CarePartners based on their relationship and communication characteristics, we chose to use latent class analysis (LCA) to identify groups of patient-CarePartner dyads. LCA is a statistical modeling approach that is useful for identifying unobserved or latent groups within a population from a set of observed characteristics based on patterns of responses (McCutcheon, 1987). It has been used in other studies to understand how combinations of behaviors or risk factors cluster together and relate to outcomes (Brantley, Kerrigan, German, Lim, & Sherman, 2017; Cleland, Lanza, Vasilenko, & Gwadz, 2017; Robinson, Knowlton, Gielen, & Gallo, 2016). Using LCA in this study allowed us to include a variety of individual items that reflect aspects of relationships and communication identified as important in earlier studies (e.g., closeness, negative interactions, difficulty communicating) and those we anticipated would be important in distinguishing dyadic relationships (e.g., patient and CarePartner attitudes about discussing HF). We included 3 dichotomous variables to represent relationship characteristics and 11 dichotomous variables to represent communication characteristics. We used LCA as an exploratory method and did not test specific hypotheses about the difference in groups’ relationship or communication and therefore did not calculate p-values. After identifying latent relationship groups, we assessed the agreement between patients’ and their CarePartners’ ratings of patient’s depressive symptoms.

We selected the best latent class model based on the Bayesian Information Criteria (BIC) (McCutcheon, 1987) and group size, allowing for groups with few members as long as their characteristics could be interpreted in meaningful ways. We described the groups based on the agreement between patients and CarePartners on patients’ depressive symptoms using the kappa statistic because kappa accounts for chance agreement (Maclure & Willett, 1987). However, because using a dichotomous cutpoint from the continuous CES-D score can artificially inflate the Kappa statistic, we also calculated the pairwise Pearson correlation between patients’ and CarePartners’ scores (Maclure & Willett, 1987).

Because of the large number of missing CarePartner ratings of patient depression, we compared the characteristics of included and excluded dyads to identify potential bias associated with missing data using a t-test for continuous variables or a chi-square test for categorical variables.

We conducted all analyses using Stata 13.0 (College Station, TX) and used the LCA Stata plugin from The Methodology Center at Pennsylvania State University (Pennsylvania State University, 2015; Lanza, Dziak, Huang, Wagner, & Collins, 2015; Lanza, Collins, Lemmon, & Schafer, 2007). The Ann Arbor VA Human Subjects Committee and the University of Michigan Institutional Review Board reviewed and approved the study.

Results

Most patients were white men who were married and no longer working and were, on average, age 68 (SD=10.4; Table 1). Their mean CES-D score was 10.1 (SD=6.1) and 51% were classified as having depressive symptoms. CarePartners were mostly white women who were married and working, with an average age of 46 (SD=12.9). Most CarePartners and patients talked to and saw one another at least once per week. There were no statistically significant differences between dyads that were included and excluded from the study (Table 1).

Table 1.

Characteristics of included and excludeda patients and CarePartners.

Variable Category Included N=201 dyads Excluded N=168 dyads p-valueb
Patient characteristics
Age, mean (SD) Years 68.3 (10.4) 67.4 (10.0) 0.41
Gender, % Female 1 1 0.87
Race/ethnicity, % White only 76 78 0.71
Black/African American only 22 19
Other/multiple 2 3
Education, % High school or less 50 49 0.94
Marital status, % Married 60 58 0.72
Employment status, % Employed 14 10 0.24
Self-reported CES-D 10 score (SD) Mean (SD) 10.1 (6.1) 9.5 (5.9) 0.41
Classified as depressed (≥10), % 51 49 0.19
CarePartner characteristics
Age, mean (SD) Years 46.3 (12.9) 48.0 (13.5) 0.21
Gender, % Female 69 61 0.10
Race/ethnicity, % White only 75 81 0.10
Black/African American only 23 19
Other/multiple 2 0
Education, % High school or less 25 32 0.19
Marital status, % Married 68 70 0.65
Employment status, % Employed 66 58 0.10
Self-reported CES-D 10 score Mean (SD) 4.6 (4.6) 4.8 (4.5) 0.57
Classified as depressed (≥10), % 12 13 0.80
Patient-CarePartner characteristics
Contact frequency: talking, % ≤Once/month 5 8 0.78
Biweekly 5 5
Weekly 25 25
More than weekly 64 63
Contact frequency: visiting, % ≤Once/month 20 24 0.79
Biweekly 16 14
Weekly 7 7
More than weekly 57 55
Distance between patient and CP, mean (SD) Miles 134 (364) 162 (477) 0.53
Open in a new tab
a

Dyads were excluded if the CarePartner did not rate the patient’s depressive symptoms at baseline.

b

P-value for difference between included and excluded dyads based on a t-test (continuous variables) or chi-square test (categorical variables).

Latent class characteristics

We identified four distinct groups using LCA based on relationship and communication dimensions (Table 2). Posterior probabilities for group assignment ranged from 0.87–0.90, indicating good model fit (Supplemental Table A). Half of dyads (n=102, 51%) were classified as Collaborative. Collaborative dyads tended to be have a close relationship and talk to and see one another frequently (Table 3). Patients in these dyads generally felt it was important to talk with their CarePartner about HF, and neither patients nor CarePartners found these discussions difficult, depressing, or confusing, despite patients sometimes experiencing negative emotions and CarePartners reporting some strain associated with helping the patients.

Table 2.

Characteristics of patients and CarePartners by assigned group from latent class analysis model.

Patient and CarePartner Characteristics Group 1: Collaborative (n=102) Group 2: Avoidant (n=33) Group 3: Distant (n=35) Group 4: Antagonistic (n=31)
Variables included in LCA Model
Close relationship (patient-reported) 97% 88% 60% 77%
Patient usually feels at least 1 negative emotion when talking to CP 38% 42% 29% 90%
CP reports regular strain in at least 1 area 15% 30% 6% 26%
Important to talk with CP about HF 95% 55% 71% 87%
Would like to talk more with CP about HF 47% 67% 49% 65%
Difficult to talk with CP about HF 6% 47% 0% 13%
Patient doesn’t want to be asked questions about HF 3% 36% 0% 29%
Patient mostly/always is frustrated when discussing HF 0% 6% 3% 45%
Patient mostly/always finds discussing HF depressing 0% 18% 0% 45%
CP doesn’t want to ask too many questions about HF 2% 18% 0% 6%
CP mostly/always is frustrated when discussing HF 0% 3% 0% 19%
CP mostly/always finds discussing HF depressing 2% 9% 0% 23%
Talk ≥2x/week (patient-reported) 93% 61% 0% 84%
See in person ≥2x/week (patient-reported) 76% 27% 11% 84%
Other Patient Characteristics
Male 99% 97% 100% 100%
Married 58% 64% 66% 55%
Employed 14% 18% 20% 3%
Depressed (CES-D ≥10) 46% 61% 34% 74%
Other CarePartner Characteristics
Male 31% 41% 23% 32%
Married 70% 78% 62% 58%
Employed 30% 36% 23% 58%
Depressed (CES-D ≥10) 12% 15% 6% 16%
CP is patient’s child 59% 73% 63% 65%
CP is another relative of patient 20% 21% 29% 19%
CP is not related to patient 24% 6% 9% 16%
Hours/week providing care to patient, mean (SD) 9.0 (20) 4.7 (17) 1.8 (3) 11.0 (19)
Hours/week providing phone support to patient, mean (SD) 6.2 (19) 2.8 (5) 6.0 (17) 3.3 (4)
Open in a new tab

Table 3.

Summary of the relative frequency of items within each group with a focus on items that distinguished groups from one another.

Patient and CarePartner Characteristics Collaborative Avoidant Distant Antagonistic
Close relationship (P) high high
Negative emotion(s) when talking (P) high
CP experiences caregiver strain high high
Important to talk with CP about HF (P) high low high
Would like to talk more with CP about HF (P)
Difficult to talk with CP about HF (P) high
Doesn’t want to be asked questions about HF (P) high high
Frustrated when discussing HF (P) low high
CP is frustrated when discussing HF high
Finds discussing HF depressing (P) low low high
CP finds discussing HF depressing low low high
CP doesn’t want to ask too many questions about HF high
Talk ≥2x/week (P) high low high
See in person ≥2x/week (P) high low  low high
Open in a new tab

“high” indicates relatively high frequency of item within group

“low” indicates relatively low frequency of item within group

CP: CarePartner

(P): patient-reported

The remainder of the dyads were divided into three similarly-sized groups: Avoidant, Distant, and Antagonistic. These groups were characterized by poorer relationship, poorer communication, or both when compared to the Collaborative group. Patients and CarePartners in Avoidant dyads (n=33, 16%) talked with average frequency (most talked twice per week or more) but saw one another in person less often than Collaborative dyads. These patients tended to find discussing HF with their CarePartner difficult and preferred to avoid these conversations, while their CarePartners tried not to ask many questions about HF and reported relatively high levels of strain associated with helping the patient. In Distant dyads (n=35, 17%), neither patients nor CarePartners reported difficulty or negative emotions associated with discussing HF, but they did not discuss HF often and were not in frequent contact. Finally, patients and CarePartners in Antagonistic dyads were frequently in contact, both by phone and in person, but patients reported high levels of negative emotions when talking to their CarePartner, including not wanting to talk about their HF and feeling frustrated or depressed when they did. CarePartners also reported relatively high levels of caregiver strain and were frequently frustrated or depressed when talking with the patient about HF.

Patients in Antagonistic dyads were less likely to be employed (3%) than patients in other groups (14–20%), while their CarePartners were more likely to be employed (58% versus 23–36%; Table 2). They also had the highest prevalence of depression (74%). Across all groups, CarePartners tended to be the patient’s adult child (59–73%). One quarter (24%) of Collaborative CarePartners were not related to the patient, compared to 6–16% of CarePartners in other relationship groups. CarePartners in Collaborative and Antagonistic groups provided the most support to patients at baseline (15 hours in each group), with Antagonistic CarePartners providing more in-person care (11 hours versus 9 hours). In both the Avoidant and Distant groups, CarePartners provided about 8 hours per week of care on average, with Distant CarePartners providing most of the care by phone (6 hours) and Avoidant CarePartners providing most of the care in person (5 hours).

We performed a post hoc sensitivity analysis to evaluate whether adjusting the definition of a close relationship influenced our results. Specifically, we classified people with a score at or above the median of 9 as having a close relationship and re-ran our LCA model. The sensitivity analysis model had a less optimal fit based on the posterior probability of group assignment, particularly for the Avoidant and Distant groups (Supplemental Table B). We therefore chose to retain the original definition and model.

Depressive Symptoms

The prevalence of depressive symptoms among patients varied across the four groups: 34% of Distant, 46% of Collaborative, 61% of Avoidant, and 74% of Antagonistic patients had depressive symptoms (CESD ≥10). Kappa statistics indicated CarePartner ratings of the presence of patient depression were moderately similar to patients’ own self-assessments overall (κ=0.18). However, observed agreement between CarePartners and the patients varied substantially across groups (Table 4). Agreement was highest in the Distant group (κ=0.44) and fair in the Collaborative group (κ=0.19), but poor in the Avoidant (κ= –0.20) and Antagonistic groups (κ= –0.01). A similar pattern of findings emerged when considering correlations between the continuous CES-D scores. Specifically, dyads in the Distant (r=0.39) and Collaborative (r=0.32) groups had moderately similar ratings of patient depressive symptoms while dyads in the Avoidant (r=0.17) and Antagonistic (r=0.004) groups had fair to poor agreement.

Table 4.

Kappa statistic for agreement between patient and CarePartner classification of patient’s depression and correlation values for similarity between patient and CarePartner ratings of patient depression using the CES-D-10.

Groupa Kappa statistic for agreement that patient CESD-10 score≥10 Correlation between patient and CarePartner CESD-10 score
Group 1: Collaborative (n=102) 0.19 0.32
Group 2: Avoidant (n=33) −0.20 0.17
Group 3: Distant (n=35) 0.44 0.39
Group 4: Antagonistic (n=31) −0.01 0.004
Open in a new tab
a

Group assignment is based on latent class analysis model incorporating 14 dichotomous variables reflecting patient and CarePartner relationship and communication.

Discussion

We identified four groups of patient-CarePartner dyads and found considerable variability in relationship and communication characteristics within these groups. We also found that half of the sample had generally positive, frequent interactions, but the other half had either frequent but negative interactions or infrequent contact. This was particularly striking given that these data came from a trial where the patient explicitly involved a CarePartner who was willing and able to be their support person. The variability in relationship and communication characteristics across groups was reflected in the degree of agreement versus disagreement on patients’ depressive symptoms. Overall, CarePartners’ reports of patients’ depressive symptoms were only moderately similar to patients’ own reports. Groups that were characterized by more positive interactions or fewer negative interactions (Collaborative or Distant) had the most agreement, where those with more negative interactions (Avoidant and Antagonistic) had least agreement.

These results are consistent with the small number of previous studies that have investigated the role of patient-caregiver relationship and communication characteristics in chronic disease management (López, López-Arrieta, & Crespo, 2005; Mayberry & Osborn, 2014; Retrum et al., 2013; Rosland et al., 2010, 2012; Spanier, 1976), and with studies of couples that have found negative interactions like criticism and avoidance are associated with negative patient outcomes (Manne, 1999; Vitaliano et al., 1993). For example, Mayberry and colleagues found that obstructive family member behaviors such as nagging and criticizing were associated with fewer self-care behaviors and poorer disease control in patients with diabetes (Mayberry & Osborn, 2014). Given the limited research on HF specifically, additional research is needed to confirm our findings and to evaluate whether interventions designed to improve relationship and communication between patients and CarePartners could also improve symptom recognition and, ultimately, depression management in patients with HF.

Patients’ depressive symptoms may influence their approach to communicating with informal caregivers, including having difficulty regulating their emotions, withdrawing from support, or needing excessive reinforcement (Detweiler-Bedell, Friedman, Leventhal, Miller, & Leventhal, 2008; Janevic et al., 2012). The difficulties patients with HF and depression have in communicating with caregivers may create barriers to caregivers understanding patient experiences and providing appropriate support. Indeed, Janevic et al. found that informal caregivers who supported patients with at least one chronic illness and comorbid depression found it harder to support the patients’ self-care activities (Janevic et al., 2012). And Pruchno et al. found that agreement between patients and caregivers about problem behaviors was associated with lower burden and depression among caregivers (Pruchno, Burant, & Peters, 1997).

Our study suggests that depression is associated with patient-CarePartner interactions and also with poorer recognition of depressive symptoms. It also suggests that reducing negative interactions, improving relationships, or enhancing communication between patients and their out-of-home CarePartners might result in caregivers better recognizing patients’ depressive symptoms. In a recent feasibility study by Trivedi et al., the team found that the relationship-focused intervention, which included 17 HF patients and their spousal caregivers, improved relationship quality and communication along with HF self-care (Trivedi et al., 2016). Taken together, these studies suggest it is possible that with more positive interaction, out-of-home caregivers can be sensitized to patients’ symptom burden.

Strengths of this study include incorporating relationship and communication ratings from both patients and CarePartners and using data across relationship and communication constructs. However, limitations include a lack of previously validated measures of relationship characteristics, and the possibility that we did not include all relevant aspects of relationship or communication. Additionally, the time frames on many of our variables differed from the time frame used to evaluate depressive symptoms (e.g., communication frequency and closeness over the past 6 months, negative emotions over the past 3 months, and depressive symptoms in the past 4 weeks). Therefore, we assumed the attributes accurately represented the past 4 weeks, which may not have been true and resulted in some measurement error, possibly misclassifying dyads in the improper latent class. Also, we used LCA as an exploratory approach, and as such, results depend somewhat upon which variables are included in the model. We excluded a substantial portion of dyads because the CarePartners did not rate the patient’s depressive symptoms. Although included and excluded dyads were similar in terms of their measured characteristics, it is reasonable to expect that CarePartners who did not rate the patient’s depressive symptoms felt less capable of rating their symptoms or were less certain of their responses. Therefore, agreement may have been lower if all study dyads had been included and additional groups may have been identified. Finally, we cannot be certain the source study sample was representative of heart failure patients in the clinics from which patients were sampled because of Human Subjects Committee restrictions on assessing patients who did not consent to participate. Regardless, the study included mostly male veterans and therefore may not have captured the full range of groups of patients and CarePartners in the broader population. Future studies using consistent time frames, valid and reliable measures, and women and non-veterans are needed to confirm these findings and identify other latent dyad groups.

Conclusion

This study is an important step toward better understanding how patients and CarePartners relate and how relationship and communication characteristics impact CarePartners’ ability to accurately assess patient experiences like depression. The results suggest it is possible that with more positive interaction out-of-home caregivers can be sensitized to patients’ symptom burden. CarePartners who recognize patients’ depressive symptoms may be able to better support their self-care activities and encourage patients to seek appropriate depression treatment. Therefore, this information could help to create more tailored programs for HF patients and their caregivers to improve this support and ultimately depression management in HF.

Supplementary Material

1

Acknowledgments

Funding sources: This original study was funded by VA Health Services Research and Development Program (HSR&D) grant #IIR 07–185. Additional financial support came from grant number P30DK092926 from the National Institute of Diabetes, Digestive and Kidney Diseases. Dr. Bouldin’s time was supported in part by a VA HSR&D Post-Doctoral Fellowship. Dr. Trivedi was supported by a VA HSR&D Career Development Award. Dr. Piette was supported by a VA HSR&D Senior Research Career Scientist Award.

Footnotes

Disclosure of Interest: The authors have no conflicts of interest to report.

Disclaimer: The views expressed in this manuscript are those of the authors and do not necessarily reflect the position or policy of the Department of Veterans Affairs or the United States Government.

References

  1. Agarwal SK, Wruck L, Quibrera M, Matsushita K, Loehr LR, Chang PP, … Coresh J (2016). Temporal Trends in Hospitalization for Acute Decompensated Heart Failure in the United States, 1998–2011. American Journal of Epidemiology, 183(5), 462–470. 10.1093/aje/kwv455 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Andresen EM, Malmgren JA, Carter WB, & Patrick DL (1994). Screening for depression in well older adults: evaluation of a short form of the CES-D (Center for Epidemiologic Studies Depression Scale). American Journal of Preventive Medicine, 10(2), 77–84. [PubMed] [Google Scholar]
  3. Artinian NT, Magnan M, Sloan M, & Lange MP (2002). Self-care behaviors among patients with heart failure. Heart & Lung: The Journal of Critical Care, 31(3), 161–172. [DOI] [PubMed] [Google Scholar]
  4. Blecker S, Paul M, Taksler G, Ogedegbe G, & Katz S (2013). Heart Failure Associated Hospitalizations in the United States. Journal of the American College of Cardiology, 61(12). 10.1016/j.jacc.2012.12.038 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Brantley ML, Kerrigan D, German D, Lim S, & Sherman SG (2017). Identifying Patterns of Social and Economic Hardship Among Structurally Vulnerable Women: A Latent Class Analysis of HIV/STI Risk. AIDS and Behavior. 10.1007/s10461-017-1673-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Buck HG, Mogle J, Riegel B, McMillan S, & Bakitas M (2015). Exploring the Relationship of Patient and Informal Caregiver Characteristics with Heart Failure Self-Care Using the Actor-Partner Interdependence Model: Implications for Outpatient Palliative Care. Journal of Palliative Medicine, 18(12), 1026–1032. 10.1089/jpm.2015.0086 [DOI] [PubMed] [Google Scholar]
  7. Carpenter JS, Andrykowski MA, Wilson J, Hall LA, Rayens MK, Sachs B, & Cunningham LL (1998). Psychometrics for two short forms of the Center for Epidemiologic Studies-Depression Scale. Issues in Mental Health Nursing, 19(5), 481–494. [DOI] [PubMed] [Google Scholar]
  8. Chang L-Y, Wu S-Y, Chiang C-E, & Tsai P-S (2017). Depression and self-care maintenance in patients with heart failure: A moderated mediation model of self-care confidence and resilience. European Journal of Cardiovascular Nursing: Journal of the Working Group on Cardiovascular Nursing of the European Society of Cardiology, 16(5), 435–443. 10.1177/1474515116687179 [DOI] [PubMed] [Google Scholar]
  9. Cleland CM, Lanza ST, Vasilenko SA, & Gwadz M (2017). Syndemic Risk Classes and Substance Use Problems among Adults in High-Risk Urban Areas: A Latent Class Analysis. Frontiers in Public Health, 5, 237 10.3389/fpubh.2017.00237 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Detweiler-Bedell JB, Friedman MA, Leventhal H, Miller IW, & Leventhal EA (2008). Integrating co-morbid depression and chronic physical disease management: identifying and resolving failures in self-regulation. Clinical Psychology Review, 28(8), 1426–1446. 10.1016/j.cpr.2008.09.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Graven LJ, Grant JS, Vance DE, Pryor ER, Grubbs L, & Karioth S (2015). Predicting depressive symptoms and self-care in patients with heart failure. American Journal of Health Behavior, 39(1), 77–87. 10.5993/AJHB.39.1.9 [DOI] [PubMed] [Google Scholar]
  12. Irwin M, Artin KH, & Oxman MN (1999). Screening for depression in the older adult: criterion validity of the 10-item Center for Epidemiological Studies Depression Scale (CES-D). Archives of Internal Medicine, 159(15), 1701–1704. [DOI] [PubMed] [Google Scholar]
  13. Janevic MR, Rosland A-M, Wiitala W, Connell CM, & Piette JD (2012). Providing support to relatives and friends managing both chronic physical illness and depression: the views of a national sample of U.S. adults. Patient Education and Counseling, 89(1), 191–198. 10.1016/j.pec.2012.05.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Jani BD, Mair FS, Roger VL, Weston SA, Jiang R, & Chamberlain AM (2016). Comorbid Depression and Heart Failure: A Community Cohort Study. PloS One, 11(6), e0158570 10.1371/journal.pone.0158570 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Joo H, Fang J, Losby JL, & Wang G (2015). Cost of informal caregiving for patients with heart failure. American Heart Journal, 169(1), 142–148.e2. 10.1016/j.ahj.2014.10.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Kohout FJ, Berkman LF, Evans DA, & Cornoni-Huntley J (1993). Two shorter forms of the CES-D (Center for Epidemiological Studies Depression) depression symptoms index. Journal of Aging and Health, 5(2), 179–193. [DOI] [PubMed] [Google Scholar]
  17. Lanza ST, Collins LM, Lemmon DR, & Schafer JL (2007). PROC LCA: A SAS Procedure for Latent Class Analysis. Structural Equation Modeling: A Multidisciplinary Journal, 14(4), 671–694. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Lanza ST, Dziak J, Huang L, Wagner A, & Collins LM (2015). LCA Stata plugin users’ guide (Version 1.2). The Pennsylvania State University; Retrieved from https://methodology.psu.edu/downloads/lcastata [Google Scholar]
  19. Lee KS, Lennie TA, Yoon JY, Wu J-R, & Moser DK (2017). Living Arrangements Modify the Relationship Between Depressive Symptoms and Self-care in Patients With Heart Failure. The Journal of Cardiovascular Nursing, 32(2), 171–179. 10.1097/JCN.0000000000000327 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Levy D, Kenchaiah S, Larson MG, Benjamin EJ, Kupka MJ, Ho KKL, … Vasan RS (2002). Long-term trends in the incidence of and survival with heart failure. The New England Journal of Medicine, 347(18), 1397–1402. 10.1056/NEJMoa020265 [DOI] [PubMed] [Google Scholar]
  21. López J, López-Arrieta J, & Crespo M (2005). Factors associated with the positive impact of caring for elderly and dependent relatives. Archives of Gerontology and Geriatrics, 41(1), 81–94. 10.1016/j.archger.2004.12.001 [DOI] [PubMed] [Google Scholar]
  22. Maclure M, & Willett WC (1987). Misinterpretation and misuse of the kappa statistic. American Journal of Epidemiology, 126(2), 161–169. [DOI] [PubMed] [Google Scholar]
  23. Manne SL (1999). Intrusive thoughts and psychological distress among cancer patients: the role of spouse avoidance and criticism. Journal of Consulting and Clinical Psychology, 67(4), 539–546. [DOI] [PubMed] [Google Scholar]
  24. Mayberry LS, & Osborn CY (2014). Family involvement is helpful and harmful to patients’ self-care and glycemic control. Patient Education and Counseling, 97(3), 418–425. 10.1016/j.pec.2014.09.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. McAlister FA, Stewart S, Ferrua S, & McMurray JJJV (2004). Multidisciplinary strategies for the management of heart failure patients at high risk for admission: a systematic review of randomized trials. Journal of the American College of Cardiology, 44(4), 810–819. 10.1016/j.jacc.2004.05.055 [DOI] [PubMed] [Google Scholar]
  26. McCutcheon AL (1987). Latent Class Analysis. Newbury Park, CA: Sage. [Google Scholar]
  27. Moser DK, & Watkins JF (2008). Conceptualizing self-care in heart failure: a life course model of patient characteristics. The Journal of Cardiovascular Nursing, 23(3), 205–218; quiz 219–220. 10.1097/01.JCN.0000305097.09710.a5 [DOI] [PubMed] [Google Scholar]
  28. Mozaffarian D, Go AS, Arnett DK, Blaha MJ, Cushman M, Das SR, Stroke Statistics Subcommittee. (2016). Heart Disease and Stroke Statistics-2016 Update: A Report From the American Heart Association. Circulation, 133(4), e38–60. 10.1161/CIR.0000000000000350 [DOI] [PubMed] [Google Scholar]
  29. Neumann PJ, Araki SS, & Gutterman EM (2000). The use of proxy respondents in studies of older adults: lessons, challenges, and opportunities. Journal of the American Geriatrics Society, 48(12), 1646–1654. [DOI] [PubMed] [Google Scholar]
  30. Pennsylvania State University, T. M. C. (2015). LCA Stata Plugin (Version 1.2) (Version 1.2) [Stata; ]. University Park: Retrieved from https://methodology.psu.edu/downloads/lcastata [Google Scholar]
  31. Piette JD, Gregor MA, Share D, Heisler M, Bernstein SJ, Koelling T, & Chan P (2008). Improving heart failure self-management support by actively engaging out-of-home caregivers: results of a feasibility study. Congestive Heart Failure (Greenwich, Conn.), 14(1), 12–18. [DOI] [PubMed] [Google Scholar]
  32. Piette JD, Striplin D, Marinec N, Chen J, & Aikens JE (2015). A randomized trial of mobile health support for heart failure patients and their informal caregivers: impacts on caregiver-reported outcomes. Medical Care, 53(8), 692–699. 10.1097/MLR.0000000000000378 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Piette JD, Striplin D, Marinec N, Chen J, Trivedi RB, Aron DC, … Aikens JE (2015). A Mobile Health Intervention Supporting Heart Failure Patients and Their Informal Caregivers: A Randomized Comparative Effectiveness Trial. Journal of Medical Internet Research, 17(6), e142 10.2196/jmir.4550 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Pruchno RA, Burant CJ, & Peters ND (1997). Understanding the well-being of care receivers. The Gerontologist, 37(1), 102–109. [DOI] [PubMed] [Google Scholar]
  35. Quinn C, Dunbar SB, & Higgins M (2010). Heart failure symptom assessment and management: can caregivers serve as proxy? The Journal of Cardiovascular Nursing, 25(2), 142–148. 10.1097/JCN.0b013e3181bf93a0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Retrum JH, Nowels CT, & Bekelman DB (2013). Patient and caregiver congruence: the importance of dyads in heart failure care. The Journal of Cardiovascular Nursing, 28(2), 129–136. 10.1097/JCN.0b013e3182435f27 [DOI] [PubMed] [Google Scholar]
  37. Riegel B, Moser DK, Anker SD, Appel LJ, Dunbar SB, Grady KL, … American Heart Association Interdisciplinary Council on Quality of Care and Outcomes Research. (2009). State of the science: promoting self-care in persons with heart failure: a scientific statement from the American Heart Association. Circulation, 120(12), 1141–1163. 10.1161/CIRCULATIONAHA.109.192628 [DOI] [PubMed] [Google Scholar]
  38. Robinson AC, Knowlton AR, Gielen AC, & Gallo JJ (2016). Substance use, mental illness, and familial conflict non-negotiation among HIV-positive African-Americans: latent class regression and a new syndemic framework. Journal of Behavioral Medicine, 39(1), 1–12. 10.1007/s10865-015-9670-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Rosland A-M, Heisler M, Choi H-J, Silveira MJ, & Piette JD (2010). Family influences on self-management among functionally independent adults with diabetes or heart failure: do family members hinder as much as they help? Chronic Illness, 6(1), 22–33. 10.1177/1742395309354608 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Rosland A-M, Heisler M, & Piette JD (2012). The impact of family behaviors and communication patterns on chronic illness outcomes: a systematic review. Journal of Behavioral Medicine, 35(2), 221–239. 10.1007/s10865-011-9354-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Rutledge T, Reis VA, Linke SE, Greenberg BH, & Mills PJ (2006). Depression in heart failure a meta-analytic review of prevalence, intervention effects, and associations with clinical outcomes. Journal of the American College of Cardiology, 48(8), 1527–1537. 10.1016/j.jacc.2006.06.055 [DOI] [PubMed] [Google Scholar]
  42. Sayers SL, Riegel B, Pawlowski S, Coyne JC, & Samaha FF (2008). Social support and self-care of patients with heart failure. Annals of Behavioral Medicine: A Publication of the Society of Behavioral Medicine, 35(1), 70–79. 10.1007/s12160-007-9003-x [DOI] [PubMed] [Google Scholar]
  43. Spanier GB (1976). Measuring Dyadic Adjustment: New Scales for Assessing the Quality of Marriage and Similar Dyads. Journal of Marriage and Family, 38(1), 15–28. 10.2307/350547 [DOI] [Google Scholar]
  44. Strömberg A (2013). The situation of caregivers in heart failure and their role in improving patient outcomes. Current Heart Failure Reports, 10(3), 270–275. 10.1007/s11897-013-0146-8 [DOI] [PubMed] [Google Scholar]
  45. Talley RC, & Crews JE (2007). Framing the public health of caregiving. American Journal of Public Health, 97(2), 224–228. 10.2105/AJPH.2004.059337 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Thornton M, & Travis SS (2003). Analysis of the reliability of the modified caregiver strain index. The Journals of Gerontology. Series B, Psychological Sciences and Social Sciences, 58(2), S127–132. [DOI] [PubMed] [Google Scholar]
  47. Trivedi R, Slightam C, Fan VS, Rosland A-M, Nelson K, Timko C, … Piette JD (2016). A Couples’ Based Self-Management Program for Heart Failure: Results of a Feasibility Study. Frontiers in Public Health, 4, 171 10.3389/fpubh.2016.00171 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Vitaliano PP, Young HM, Russo J, Romano J, & Magana-Amato A (1993). Does expressed emotion in spouses predict subsequent problems among care recipients with Alzheimer’s disease? Journal of Gerontology, 48(4), P202–209. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS1512329-supplement-1.pdf (95.8KB, pdf)

RESOURCES