Abstract
Background:
Family caregivers of cancer patients are very involved in communication with healthcare teams; however, little is known about their experiences. Limited information is known about how the type of cancer patients have impact caregiving experiences.
Objectives:
This study seeks to compare the caregiving experience of caregivers of hospice lung cancer patients with hospice caregivers of patients with all other cancer types.
Method:
This study is based on a secondary analysis of data generated from a parent study evaluating a behavioral intervention with caregivers of hospice cancer patients.
Results:
When comparing caregiving experiences by patient diagnosis, significant differences were found in caregivers of hospice lung cancer demographics and experiences with caregiver centered communication. Specifically, caregivers of non-lung cancer patients has significantly more trouble with exchange of information, fostering relationships, and decision making with their hospice team.
Conclusion:
More research is needed to understand the impact of lung cancer on caregiver centered communication and the necessary interventions required to address these issues.
Lay summary:
This project compared experiences of family caregivers of hospice lung cancer with experiences of family caregivers of patients with other kinds of cancer. We found at caregivers of hospice lung cancer patients have lower caregiver centered communication than other types of cancer. These data suggest a need to understand the unique needs and experiences of caregivers of family lung cancer patients.
Caregiving for adult cancer patients is filled with unexpected detours and changes over time, as the patient’s clinical course changes. Family caregiver burden is influenced by interactions between the physical symptoms and problems of both the patient and caregiver.1 The more advanced the cancer, the greater the stressors on caregivers, who typically are untrained to manage increasing and rapidly changing physical and psychological challenges.2 Caregivers of patients across the cancer trajectory are at increased risk for physical and mental morbidity with a psychological burden often greater than that of the patient.2 3
Research to date has described the characteristics and burdens associated with caregivers of oncology patients. Family caregivers of patients with advanced cancer are most often female, a spouse or adult child, and 50 to 65 years of age.4,5 Three studies found that up to half of family caregivers of patients with advanced cancer continue to work outside the home but eventually need to reduce their hours or stop working.5,6 The overall mental health of such caregivers falls below the 30th percentile of the general US population, with an increased incidence of both major depressive episodes and generalized anxiety disorder.5,7 A recent review of interventions for caregivers identified increased risks of poor mental health outcomes, symptom distress, poor prognostic understanding, and adverse bereavement outcomes. They found moderate to severe levels of depression (39%) and anxiety (47%).8 It was also noted that several sociodemographic variables influence depression and anxiety; such as age, relationship, and gender.8
Caregiving and Health Care Provider Communication
As caregivers play a critical role in managing and assisting with a patient’s care plan, they are called upon to frequently communicate with health care providers in this process. A NCI monograph on patient-centered communication noted that there has been limited attention paid to assessing communication between family members and the health care delivery team.9 Family caregivers of cancer patients are very involved in communication with healthcare teams; however, little is known about their experiences. Our preliminary work found that caregivers value communication with health care providers who are attentive, genuine, focused on patient and caregiver experiences, sensitive to unmet information needs, and responsive to different communication preferences for patients and their caregivers.10
Caregivers of Lung Cancer Patients
A systematic review of psychosocial interventions found that communications interventions targeting caregivers of lung cancer patients led to improvements in burden, anxiety, depression, and overall quality of life.11 Few studies have been conducted related specifically to caregivers of advanced lung cancer patients; however, the small body of existing evidence has shown that caregivers of lung cancer patients are more negatively impacted by the patient’s quality of life than the stage of the disease and poor management of patient symtoms causes increased burden leading to significant distress and lower quality of life.12,13,14 Because of the rapid progression of the disease, one study found that caregivers of lung cancer patients are particularly in need of informational support.15 More specifically, Schook et all explored why caregivers of lung cancer so often used the Internet to find information and used a online “ask physician” forum rather than communicating with their own health care providers.16 Among the reasons they found that the ability to gain information quickly and get a quick replay rather than waiting on an upcoming appointment eased their anxiety.
There is a small amount of research on hospice lung cancer patients. These findings suggest hospice utilization and length of stay were low for patients with lung cancer and they experienced several barriers to enrollment.17,18 Likewise, while their family caregivers faced different stressors there was a negative health effect on all hospice caregivers regardless of diagnosis. 19 Specific research comparing hospice caregivers by type of cancer was not be found.
A Framework for Caregiver Centered Communication
Figure 1 illustrates the conceptual model underlying caregiver-centered communication with health care providers. The model integrates ideas set forth in the NCI monograph authored by Epstein and Streets regarding patient-centered communication with Pearlin’s et al.’s Stress Process Model.9,20 These frameworks recognize that the caregiving trajectory is tied to the patient cancer trajectory and that different phases of caregiving and illness experiences introduce unique communication needs. Pearlin et al.’s Stress Process Model holds that the background and contextual factors of a caregiving situation (e.g., a person’s socioeconomic status and caregiving history) influence the primary, secondary, and intrapsychic stressors caregiver’s experience.20 These stressors result in patient and family needs that are likely to impact communication with health care providers. In their description of patient-centered communication, Epstein and Street identify six functions of patient communication with the health care team.9 The conceptual model shown in Figure 1 adapts those functions as they apply to caregiver-centered communication at the end of life. The adapted functions include building trust (fostering relationships), managing care, exchanging information, responding to emotions, and making decisions. These functions impact proximal outcomes (knowledge of symptom management), intermediate outcomes (social support) that in turn impact primary health outcomes of the family caregiving experience (depression and burden).9
Figure 1:
Caregiver Centered Communication Conceptual Model
Note: Modified from Epstein RM, Street RL. Patient-Centered Communication in Cancer Care: Promoting Healing and Reducing Suffering. In: Institute NC, ed. Bethesda, MD2007
Study Purpose and Research Questions
Based on the literature from several systematic reviews highlighting the unique issues for caregivers of lung cancer patients15, including their need for informational support and positive response to psychosocial interventions15 we sought to compare the caregiving experience of caregivers of hospice lung cancer patients with non-lung cancer, especially related to outcomes and communication. This paper asks: 1) How do hospice caregivers of patients with lung cancer compare with non-lung cancer? 2) How does communication with the health care team differ for caregivers of hospice lung cancer patients’ with non-lung cancer?
Methods
This study is based on a secondary analysis of data generated from a parent study called ACCESS (Access for Cancer Caregivers to Education and Support for Shared decision-making) (R01CA203999), a five-year, cluster-crossover, pragmatic, randomized clinical trial testing an intervention that provides education and emotional support to family caregivers of hospice cancer patients based on a structured process for shared decision-making.21 ACCESS had three components: 1) Facebook groups to provide education and emotional support to family caregivers, 2) web-conferencing to involve caregivers in hospice care plan meetings, and 3) a structured shared decision-making process to guide team discussion. The components were designed to facilitate communication and shared decision-making between hospice staff and family caregivers, lower caregiver anxiety, and increase caregiver knowledge regarding pain. Seven hospice agencies were randomly assigned to one of three groups in the cluster-crossover clinical trial. Participants enrolled in hospices assigned to Group 1 received usual hospice care. Participants enrolled in hospices assigned to Group 2 received usual care and participated in facilitated hidden (private) Facebook groups for informational and emotional support. Participants enrolled in hospices assigned to Group 3 receive usual care, participated in Facebook groups, and attended (through web conferencing or telephone) biweekly hospice interdisciplinary team meetings where they took part in shared decision-making. Each agency was assigned to each condition for 12 months with a 90-day washout period between each crossover to prevent contamination between conditions.
Sample
The sample of the sub-study presented here includes consenting participants of the ACCESS parent study. Participants are adult family caregivers of cancer patients enrolled in one of seven hospice agencies in one Midwestern state. The database was divided into those with a lung cancer diagnosis (includes all types and histologies including mesothelioma) and those with other cancer diagnoses for comparison purposes. The total sample size for the parent study is 489 cases. There are valid data on cancer type for 478 of these cases; however, for all variables in our linear models, there are valid data for 272 cases.
Measures
The Caregiver-Centered Communication Questionnaire (CCCQ) 22,23 is a 30-item self-report instrument that assesses the extent to which family caregivers believe their needs and perspectives are adequately acknowledged and addressed during interactions with healthcare teams. CCCQ total scores range from 30 to 150; higher scores reflect stronger perceptions of caregiver-centered communication. The CCCQ has strong evidence of reliability and validity. 22,24 Five subscales measure the 5 communication functions noted in the theoretical model: exchange of information between hospice team and caregiver, fostering health relationships between caregiver and hospice team, the perception of the hospice team recognizing and responding to emotions expressed by caregiver, managing care by caregiver, and decision-making by caregiver.
For caregiver outcomes, we examined depression measured by the PHQ-9 at baseline, last survey, and the difference between the two; anxiety measured by the GAD-7 at baseline, last survey, and the difference between the two; total quality of life measured by the Caregiver Quality of Life Index – Revised at baseline, burden measured by the Zarit Burden Interview at baseline, caregiver report of patient symptoms measured by the Edmonton Symptom Assessment System – Revised at baseline, pain knowledge measured by the Family Pain Questionnaire at baseline, pain experience measured by the Family Pain Questionnaire at baseline, and caregivers’ perceptions of Caregiver-Centered Communication with their hospice team was measured by the overall CCCQ at first follow-up as well as its 5 subscales at first follow-up. We examine overall CCCQ as the outcome of our linear models in order to determine how lung cancer type might affect perception of caregiver communication with the hospice team while adjusting for demographic and contextual variables. Measures for the contextual variables are described elsewhere in the literature.23,25–30
Our primary explanatory variable of interest is cancer type. For the purposes of the analyses and ease of interpretation, we compared caregivers of hospice patients with lung cancer against all other types of cancer caregivers. In addition, we utilize both patient and caregiver demographics to examine differences between caregivers of patients with lung cancer and caregivers of patients with other types of cancers, and to adjust for demographics in our linear models. These demographic variables include patient age, race, ethnicity, marital status, and education; and caregiver age, race, ethnicity, marital status, education, employment status, income, and relationship to patient. In addition, we adjust for contextual variables such as patient residence, days on hospice, and intervention condition.
Data analysis
To answer our research questions, we use a series of analyses starting with bivariate and proceeding to multivariate. First, we examine demographic variables for patients and caregivers. Second, we look at bivariate analyses, specifically Wilcoxon tests and chi-square tests to examine differences between patients and caregivers of patients with lung cancer and all other cancers on both demographic variables and baseline measures.
Given our focus on caregivers’ perceptions of their communication with hospice teams, we estimated linear models to adjust for demographic variables and examine the independent effect of cancer type on CCCQ scores specifically. These linear models utilize a block-wise approach. The first model contains only the outcome and explanatory variable cancer type, the second model adjusts for caregiver demographic controls, and the third model adjusts for contextual controls.
We examined both the significance and estimate for the explanatory variable in our linear models (≤.05 considered statistical evidence of a relationship between input and outcome) to determine whether we are able to reject or must retain the null hypotheses as well as other variables for any other practically significant findings. We also examined the overall model fit statistics to determine the model fit to the data. We checked assumptions and diagnostics using a scatterplot for residuals and fitted values, a normal q-q plot, and a scale-location plot. Missing data were handled using case-wise deletion.
Results
Demographic characteristics are presented for patients and caregivers in Tables 1 and 2. Among lung cancer patients, the average age is 72.5, and they are marginally more male (54%) and majority white (85%). These patients are primarily partnered (38%) or widowed (32%) and are primarily high school educated (39%) or with some college education (31%). Patients with other types of cancer are similar in all respects as shown in Table 1. Among caregivers of lung cancer patients, the average age is 58, and they are majority female (78%), white (83%), and partnered (73%). These caregivers are at least high school educated (98%), either unemployed/retired (46%) or employed full-time (48%), and a marginal majority have household incomes of over $40,000 (65%). These caregivers are typically either adult children (48%) or spouses (26%) of the patient. Again, caregivers of patients without lung cancer are largely similar in terms of demographic makeup aside from marital status, in that there are more divorced/separated caregivers of lung cancer patients than those with other types of cancer.
Table 1.
Summary of Patient Demographics by Cancer Type.
| Lung, Mesothelioma (N=103) |
Other (N=375) |
P-value1 | |
|---|---|---|---|
|
| |||
| Age | |||
| Mean (SD) | 72.5 (10.1) | 71.5 (14.4) | 0.89 |
| Median [Min, Max] | 73.0 [52.0, 95.0] | 72.0 [25.0, 98.0] | |
| Missing | 4 (3.9%) | 25 (6.7%) | |
| Gender | |||
| Male | 56 (54%) | 189 (51%) | 0.56 |
| Female | 47 (46%) | 185 (49%) | |
| Missing | 0 (0%) | 1 (0.3%) | |
| Race | |||
| Black/AA | 12 (12%) | 42 (11%) | 0.22 |
| White/Caucasian | 85 (85%) | 326 (88%) | |
| Other | 3 (3%) | 3 (1%) | |
| Missing | 3 (2.9%) | 4 (1.1%) | |
| Ethnicity | |||
| Non-hispanic | 99 (96%) | 368 (99%) | 0.13 |
| Hispanic | 4 (4%) | 4 (1%) | |
| Missing | 0 (0%) | 3 (0.8%) | |
| Marital Status | |||
| Single, never partnered | 14 (14%) | 39 (11%) | 0.46 |
| Partnered | 39 (38%) | 166 (45%) | |
| Divorced/separated | 17 (17%) | 50 (13%) | |
| Widowed | 33 (32%) | 111 (30%) | |
| Other | 0 (0%) | 5 (1%) | |
| Missing | 0 (0%) | 4 (1.1%) | |
| Education | |||
| Less than high school | 15 (15%) | 43 (12%) | 0.69 |
| High school/GED | 40 (39%) | 140 (38%) | |
| Some college/trade school | 32 (31%) | 107 (29%) | |
| Undergraduate degree | 11 (11%) | 44 (12%) | |
| Graduate/professional degree | 5 (5%) | 30 (8%) | |
| Other | 0 (0%) | 4 (1%) | |
| Missing | 0 (0%) | 7 (1.9%) | |
P values were derived from Chi-square tests for categorical variables and Wilcoxon two-sample tests for continuous variables.
Table 2.
Summary of Caregiver Demographics by Cancer Type.
| Lung, Mesothelioma (N=103) |
Other (N=375) |
P-value1 | |
|---|---|---|---|
|
| |||
| Age | |||
| Mean (SD) | 56.9 (11.9) | 56.0 (12.5) | 0.83 |
| Median [Min, Max] | 58.0 [26.0, 90.0] | 57.5 [22.0, 89.0] | |
| Missing | 2 (1.9%) | 5 (1.3%) | |
| Gender | |||
| Male | 23 (22%) | 87 (23%) | 0.95 |
| Female | 80 (78%) | 287 (77%) | |
| Missing | 0 (0%) | 1 (0.3%) | |
| Race | |||
| Black/AA | 12 (12%) | 43 (12%) | 0.22 |
| Other | 5 (5%) | 7 (2%) | |
| White/Caucasian | 84 (83%) | 321 (87%) | |
| Missing | 2 (1.9%) | 4 (1.1%) | |
| Ethnicity | |||
| Non-hispanic | 99 (96%) | 365 (98%) | 0.30 |
| Hispanic | 4 (4%) | 6 (2%) | |
| Missing | 0 (0%) | 4 (1.1%) | |
| Marital Status | |||
| Single, never partnered | 7 (7%) | 59 (16%) | 0.016 |
| Partnered | 75 (73%) | 270 (72%) | |
| Divorced/separated | 15 (15%) | 26 (7%) | |
| Widowed | 3 (3%) | 14 (4%) | |
| Other | 3 (3%) | 4 (1%) | |
| Missing | 0 (0%) | 2 (0.5%) | |
| Education | |||
| Less than high school | 2 (2%) | 11 (3%) | 0.17 |
| High school/GED | 26 (25%) | 72 (19%) | |
| Some college/trade school | 46 (45%) | 137 (37%) | |
| Undergraduate degree | 19 (18%) | 80 (22%) | |
| Graduate/professional degree | 10 (10%) | 65 (18%) | |
| Other | 0 (0%) | 5 (1%) | |
| Missing | 0 (0%) | 5 (1.3%) | |
| Employment Status | |||
| Not employed/retired | 46 (46%) | 158 (43%) | 0.14 |
| Employed part-time | 5 (5%) | 44 (12%) | |
| Employed full-time | 48 (48%) | 167 (45%) | |
| Missing | 4 (3.9%) | 6 (1.6%) | |
| Income | |||
| Under $20,000 per year | 11 (13%) | 44 (13%) | 0.54 |
| $20,000 to $39,999 | 19 (22%) | 70 (21%) | |
| $40,000 to $69,000 | 20 (23%) | 102 (31%) | |
| Over $70,000 | 36 (42%) | 117 (35%) | |
| Missing | 17 (16.5%) | 42 (11.2%) | |
| Relationship to Patient | |||
| Adult child | 49 (48%) | 161 (43%) | 0.69 |
| Spouse/partner | 27 (26%) | 118 (32%) | |
| Other relative | 14 (14%) | 55 (15%) | |
| Other | 13 (13%) | 40 (11%) | |
| Missing | 0 (0%) | 1 (0.3%) | |
P values were derived from Chi-square tests for categorical variables and Wilcoxon two-sample tests for continuous variables.
Our bivariate analyses suggested a significant difference in mean burden (Zarit) and mean patient symptom (ESAS) scores between lung cancer caregivers and other types of cancer caregivers, with lung cancer patients having a greater mean for symptoms and their caregivers a lesser mean for burden. These results are presented in Table 3. We also examined mean differences between lung cancer caregivers and other types of cancer caregivers on the various subscales of the CCCQ. These results are presented in Table 4. They suggest lung cancer caregivers reported worse communication than other types of cancer caregivers on the exchanging information (p≤ .02), fostering relationships (p≤ .05), and decision-making (p≤ .03) subscales, with caregivers of patients with lung cancer having a lower average score on these subscales than caregivers of patients with other types of cancer. We did not find a significant difference between caregivers of patients with lung cancer and those with other types of cancer on depression, anxiety, quality of life, or pain knowledge and experience.
Table 3.
Summary of Outcomes by Cancer Type
| Lung, Mesothelioma (N=103) |
Other (N=375) |
P-value1 | |
|---|---|---|---|
|
| |||
| PHQ9 | |||
| Mean (SD) | 7.07 (6.16) | 7.90 (6.37) | 0.28 |
| Median [Min, Max] | 5.50 [0, 25.0] | 6.00 [0, 24.0] | |
| Missing | 7 (6.8%) | 22 (5.9%) | |
| Last PHQ9 | |||
| Mean (SD) | 6.46 (6.30) | 6.85 (5.57) | 0.29 |
| Median [Min, Max] | 4.00 [0, 27.0] | 6.00 [0, 26.0] | |
| Missing | 40 (38.8%) | 132 (35.2%) | |
| Difference PHQ9 | |||
| Mean (SD) | 1.03 (3.28) | 1.01 (4.77) | 0.74 |
| Median [Min, Max] | 1.00 [-11.0, 9.00] | 1.00 [-15.0, 20.0] | |
| Missing | 45 (43.7%) | 144 (38.4%) | |
| GAD7 | |||
| Mean (SD) | 7.52 (5.52) | 8.68 (5.95) | 0.12 |
| Median [Min, Max] | 7.00 [0, 21.0] | 7.00 [0, 21.0] | |
| Missing | 7 (6.8%) | 22 (5.9%) | |
| Last GAD7 | |||
| Mean (SD) | 6.49 (5.60) | 7.57 (5.30) | 0.11 |
| Median [Min, Max] | 6.00 [0, 21.0] | 7.00 [0, 21.0] | |
| Missing | 40 (38.8%) | 131 (34.9%) | |
| Difference GAD7 | |||
| Mean (SD) | 1.30 (5.06) | 1.12 (4.86) | 0.99 |
| Median [Min, Max] | 1.00 [-10.0, 14.0] | 1.00 [-13.0, 17.0] | |
| Missing | 47 (45.6%) | 142 (37.9%) | |
| Emotional Quality of Life | |||
| Mean (SD) | 6.70 (2.54) | 6.67 (2.37) | 0.83 |
| Median [Min, Max] | 7.00 [0, 10.0] | 7.00 [0, 10.0] | |
| Missing | 1 (1.0%) | 6 (1.6%) | |
| Social Quality of Life | |||
| Mean (SD) | 5.96 (2.94) | 5.78 (2.81) | 0.54 |
| Median [Min, Max] | 6.00 [0, 10.0] | 6.00 [0, 10.0] | |
| Missing | 1 (1.0%) | 5 (1.3%) | |
| Financial Quality of Life | |||
| Mean (SD) | 6.47 (2.81) | 6.19 (2.77) | 0.35 |
| Median [Min, Max] | 7.00 [0, 10.0] | 7.00 [0, 10.0] | |
| Missing | 2 (1.9%) | 13 (3.5%) | |
| Physical Quality of Life | |||
| Mean (SD) | 6.34 (2.22) | 6.26 (2.28) | 1 |
| Median [Min, Max] | 6.00 [0, 10.0] | 7.00 [0, 10.0] | |
| Missing | 1 (1.0%) | 5 (1.3%) | |
| Total Quality of Life | |||
| Mean (SD) | 25.5 (8.81) | 24.9 (8.30) | 0.59 |
| Median [Min, Max] | 26.0 [4.00, 40.0] | 25.0 [0, 40.0] | |
| Missing | 2 (1.9%) | 14 (3.7%) | |
| Zarit Interview - Burden | |||
| Mean (SD) | 8.43 (5.72) | 10.1 (6.33) | 0.02 |
| Median [Min, Max] | 7.00 [0, 23.0] | 10.0 [0, 28.0] | |
| Missing | 4 (3.9%) | 14 (3.7%) | |
| Edmonton Symptom | |||
| Mean (SD) | 50.5 (11.6) | 43.8 (15.2) | 0.03 |
| Median [Min, Max] | 49.0 [31.0, 81.0] | 43.0 [7.00, 78.0] | |
| Missing | 70 (68.0%) | 259 (69.1%) | |
| Family Pain Questionnaire - Knowledge | |||
| Mean (SD) | 45.5 (12.9) | 45.1 (12.2) | 0.98 |
| Median [Min, Max] | 45.0 [20.0, 83.0] | 45.0 [8.00, 90.0] | |
| Missing | 10 (9.7%) | 53 (14.1%) | |
| Family Pain Questionnaire - Experience | |||
| Mean (SD) | 34.3 (13.2) | 33.9 (12.9) | 0.71 |
| Median [Min, Max] | 37.0 [0, 58.0] | 36.0 [0, 61.0] | |
| Missing | 11 (10.7%) | 42 (11.2%) | |
| CCCQ Total Score | |||
| Mean (SD) | 109 (24.4) | 117 (19.7) | 0.05 |
| Median [Min, Max] | 113 [30.0, 150] | 116 [56.0, 150] | |
| Missing | 46 (44.7%) | 150 (40.0%) | |
P values were derived from Wilcoxon two-sample tests for all variables.
Table 4.
Summary of CCCQ Subscales by Cancer Type
| Lung, Mesothelioma (N=103) |
Other (N=375) |
P-value1 | |
|---|---|---|---|
|
| |||
| CCCQ: Exchanging Information | |||
| Mean (SD) | 18.8 (4.23) | 20.1 (3.52) | 0.02 |
| Median [Min, Max] | 20.0 [5.0, 25.0] | 20.0 [8.0, 25.0] | |
| Missing | 44 (42.7%) | 149 (39.7%) | |
| CCCQ: Fostering Health Relationships | |||
| Mean (SD) | 37.6 (8.37) | 40.4 (6.43) | 0.05 |
| Median [Min, Max] | 39.0 [10.0, 50.0] | 40.0 [21.0, 50.0] | |
| Missing | 46 (44.7%) | 150 (40.0%) | |
| CCCQ: Recognizing and Responding to Emotions | |||
| Mean (SD) | 17.1 (4.58) | 18.2 (3.96) | 0.22 |
| Median [Min, Max] | 18.0 [5.0, 25.0] | 18.0 [5.0, 25.0] | |
| Missing | 46 (44.7%) | 150 (40.0%) | |
| CCCQ: Managing Care | |||
| Mean (SD) | 14.5 (3.29) | 15.4 (2.79) | 0.18 |
| Median [Min, Max] | 15.0 [4.0, 20.0] | 15.0 [8.0, 20.0] | |
| Missing | 46 (44.7%) | 150 (40.0%) | |
| CCCQ: Decision-making | |||
| Mean (SD) | 20.9 (5.05) | 22.8 (4.49) | 0.03 |
| Median [Min, Max] | 21.0 [6.0, 30.0] | 23.0 [6.0, 30.0] | |
| Missing | 46 (44.7%) | 150 (40.0%) | |
P values were derived from Wilcoxon two-sample tests for all variables.
We then estimate linear models of the effect of lung cancer compared to other types of cancer on the CCCQ instrument while controlling for demographic and contextual variables. The results are presented in Table 5. In all three models, lung cancer is significantly associated with CCCQ score, indicating that caregivers of patients with lung cancer rate caregiver-centered communication 9 points less on average than caregivers of patients with any other type of cancer (b= −9.3, se = 3.016). In other words, caregivers of hospice patients with lung cancer experience worse communication with their hospice team, likely related to exchange of information, fostering relationships, and shared decision making.
Table 5.
Linear models for the relationship between cancer type and CCCQ score.
| Model 1 | Model 2 | Model 3 | |
|---|---|---|---|
|
|
|||
| Lung Cancer/Mesothelioma (ref = Other Cancer) | −8.255 ** | −9.31 ** | −8.72 ** |
| (3.070) | (3.02) | (3.12) | |
| Age (years) | 0.36 ** | 0.318 ** | |
| (0.11) | (0.11) | ||
| Race (ref = Black/African American) | |||
| White | −0.86 | −1.367 | |
| (4.33) | (4.34) | ||
| Other Race | −18.62 | −18.74 | |
| (12.31) | (12.20) | ||
| Female (ref = Male) | 2.29 | 2.54 | |
| (2.77) | (2.79) | ||
| Education (ref = Less than high school) | |||
| High school/GED | −2.75 | −1.28 | |
| (8.03) | (8.62) | ||
| Some college/trade school | −5.94 | −4.16 | |
| (7.79) | (8.40) | ||
| Undergraduate Degree | −11.86 | −9.70 | |
| (7.93) | (8.508001) | ||
| Graduate/Professional Degree | −11.141 | −7.76 | |
| (8.23) | (8.90) | ||
| Other Education | 6.38 | ||
| (20.78) | |||
| Marital Status (ref = Single, never partnered) | |||
| Partnered | −2.30 | −1.88 | |
| (4.12) | (4.08) | ||
| Widowed | −10.60 | −10.31 | |
| (8.38) | (8.30) | ||
| Divorced/separated | −8.13 | −6.39 | |
| (5.49) | (5.69) | ||
| Other Marital | 7.42 | 6.00 | |
| (10.46) | (10.31) | ||
| Caregiver Relationship (ref = Adult Child) | |||
| Spouse/Partner | 5.19 | 3.59 | |
| (3.19) | (3.41) | ||
| Other relative | 13.37 *** | 11.92 ** | |
| (3.91) | (3.95) | ||
| Other non-relative | −0.26 | 0.37 | |
| (3.91) | (3.91) | ||
| Patient Residence (ref = Private residence) | |||
| Nursing home | −2.79 | ||
| (3.46) | |||
| Other Residence | −4.42 | ||
| (5.49) | |||
| Hospice days | −0.16 | ||
| (0.26) | |||
| Study group (ref = Control) | |||
| Facebook Only | −0.54 | ||
| (3.08) | |||
| ACCESS Intervention | 1.17 | ||
| (3.18) | |||
|
|
|||
| N. obs. | 282 | 272 | 263 |
| R squared | 0.03 | 0.19 | 0.17 |
| F statistic | 7.23 | 3.60 | 2.42 |
| P value | 0.008 | 0.000 | 0.001 |
p ≤ 0.001
p ≤ 0.01
p ≤ 0.05.
Each model demonstrates a good fit to the data, but comparing the models adjusted for demographics and adjusted for demographics and contextual variables, there is no improvement to the fit in adding contextual models, so we draw our conclusions from the model adjusted for demographic characteristics only. This model explained roughly 19% of the variation in CCCQ (R2 = 0.194). This model also suggests that age is significantly associated with CCCQ as well as being a relative other than spouse or child. For age, the results suggest that as age increases, the CCCQ score also increases, and relatives other than spouse or child score, on average, 13 points higher on the CCCQ instrument.
Discussion
Generalization of these results need to be done with caution as these data are limited by a sample from one Midwestern state. Additionally, these results are specific to caregivers of hospice patients which may not represent all caregivers of patients with advanced lung cancer. Finally, these data were not collected specifically for this analysis and thus are limited by what was collected for another study.
The final model explained 19% of the variance, additional individual characteristics or hospice agency characteristics did not explain any larger variance. While all hospice family caregiver demographic variables are similar, their perceptions of the caregiver-centeredness of their communication with the hospice team is rated significantly worse than family caregivers of hospice patients with any other type of cancer. Our findings suggest that caregivers of hospice lung cancer patients have significant differences when it comes to perceptions of the caregiver-centeredness of communication with the health team as well as some caregiver outcomes when compared to caregivers of hospice patients with other types of cancer. More specifically, caregivers of lung cancer patients rated the caregiver-centeredness of communication with the health care team lower on average than caregivers of patients with other types of cancer. According to the subscale results, they perceived greater challenges in exchanging information with the health care team, fostering meaningful relationships with the team, and being actively engaged in decision making. These findings are not surprising in light of the findings by Schoonk who investigated the use of the Internet and online “ask a physician” by caregivers of lung cancer patients. That study indicated that these caregivers were anxious and in need of quick information, a second opinion, and animity in the asking of questions.16 Those findings support a premise that is supported with our data that caregivers of lung cancer patients are challenged to get information and hesitant to trust and be involved. The etiology of this difference is unclear, warranting more research.
Differences were also noted in caregiver symptoms and burden between the two groups, with a somewhat puzzling finding of greater patient symptom burden and lower overall caregiver burden for lung cancer patients and caregivers. This lower burden is in contrast with previous research which has found that the greater patient symptom burden and quicker patient decline has increased caregiver burden for those caregiving for lung cancer patients.11 Given these patients were receiving hospice care it is possible that the services of the hospice, in comparison with limited home services offered outside of hospice, may have been a reason for the reduced caregiver burden. Further research is warranted to understand this discrepancy.
This secondary analysis however has some limitations. The primary limitation is that we only had one CCCQ measure, administered 14 days after the caregiver enrolled in the original trial; thus we were unable to assess the change in CCCQ scores over the hospice enrollment period. Additionally, the study data are limited to descriptive and thus, are not causal but rather, point to the need for additional research. We also examined differences between those with complete data and those with incomplete data based on the variables in our linear model. The results (available upon request) suggested that there was a significant difference between those with complete and incomplete data on race (9% Black/AA complete to 15% Black/AA incomplete) and education (35% some college complete to 43% incomplete and 26% undergraduate degree complete to 15% undergraduate degree incomplete). This limits some of the representativeness of the valid sample as well as generalizability of results. Finally, participants were largely white indicating a racial bias in the results, communication with minority caregivers might have been different.
The caregiver-centered communication framework would suggest that less caregiver-centered communication would be reflected in significantly worse caregiver outcomes. These data found no significant differences in proximal outcomes such as pain knowledge; however, there were differences in intermediate outcomes, specifically caregiver perception of patient symptoms.
Also in conflict with the limited literature on the topic, these data found that lung cancer caregiver health outcomes, reflected in Zarit Burden Interview scores, were significantly better than other cancers. Additionally, caregiver quality of life was not significantly different, again a conflict with the literature.13,14 These are unexpected outcomes that do not align with either the previous literature or our conceptual model. Further research is required to explain this finding.
Conclusion
Our findings indicate significant differences in the caregiving experience for caregivers of hospice lung cancer patients compared with caregivers of patients with all other cancers. Additionally, some of the findings are in contrast to other studies with cancer caregivers in general. Further research is needed to understand the unique needs of hospice caregivers especially their communication needs. These findings are clinically significant as hospice teams need to take note of the increased symptom burden of lung cancer patients and the apparent higher need by caregivers for more information, a need for closer relationships, and a desire for more shared decision making as compared with caregivers of other types of cancer patients. Interventions focused on improving communication—especially exchange of information, fostering relationships, and decision making—have the potential to improve caregiver health outcomes. Continued research is needed to understand the unique communication needs of caregivers with the hospice team.
Acknowledgments
Research reported in this publication was supported by the National Cancer Institute under award number R01CA203999 (Parker Oliver). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Footnotes
Conflict of Interest: None
Contributor Information
Debra Parker Oliver, Division of Palliative Medicine, Department of Medicine, Washington University in St. Louis, Goldfarb School of Nursing, 4590 Children’s Place, Mailstop 90-29-931, St. Louis, MO. 63110.
George Demiris, Department of Biobehavioral Health Sciences, School of Nursing, Department of Biostatistics, Epidemiology and Informatics, Perelman School of Medicine, University of Pennsylvania.
Jacquelyn J. Benson, Division of Palliative Medicine, Department of Medicine, Washington University in St. Louis.
Patrick White, Division of Palliative Medicine, Department of Medicine, Washington University in St. Louis..
Audrey S. Wallace, Radiation Oncology, St. Louis Veteran Health Administration Medical Center, St. Louis, Missouri.
Kyle Pitzer, Division of Palliative Medicine, Department of Medicine, Washington University in St. Louis.
Karla T. Washington, Division of Palliative Medicine, Department of Medicine, Washington University in St. Louis.
References
- 1.Johansen S, Cvancarova M, Ruland C. The Effect of Cancer Patients’ and Their Family Caregivers’ Physical and Emotional Symptoms on Caregiver Burden. Cancer Nurs. 2018;41(2):91–99. [DOI] [PubMed] [Google Scholar]
- 2.Williams AL, McCorkle R. Cancer family caregivers during the palliative, hospice, and bereavement phases: a review of the descriptive psychosocial literature. Palliat Support Care. 2011;9(3):315–325. [DOI] [PubMed] [Google Scholar]
- 3.Nipp RD, El-Jawahri A, Fishbein JN, et al. Factors associated with depression and anxiety symptoms in family caregivers of patients with incurable cancer. Ann Oncol. 2016;27(8):1607–1612. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Washington KT, Pike KC, Demiris G, Parker Oliver D. Unique characteristics of the informal hospice cancer caregiving experience. Support Care Cancer. 2015;23:2121–2128. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Wadhwa D, Burman D, Swami N, Rodin G, Lo C, Zimmermann C. Quality of life and mental health in caregivers of outpatients with advanced cancer. Psychooncology. 2013;22(2):403–410. [DOI] [PubMed] [Google Scholar]
- 6.Hudson PL, Thomas K, Trauer T, Remedios C, Clarke D. Psychological and social profile of family caregivers on commencement of palliative care. J Pain Symptom Manage. 2011;41(3):522–534. [DOI] [PubMed] [Google Scholar]
- 7.Trevino KM, Prigerson HG, Maciejewski PK. Advanced cancer caregiving as a risk for major depressive episodes and generalized anxiety disorder. Psychooncology. 2018;27(1):243–249. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Oechsle K, Ullrich A, Marx G, et al. Psychological burden in family caregivers of patients with advanced cancer at initiation of specialist inpatient palliative care. BMC Palliat Care. 2019;18(1):102. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Epstein RM, Street RL. Patient-Centered Communication in Cancer Care: Promoting Healing and Reducing Suffering. In: Institute NC, ed. Bethesda, MD: 2007. [Google Scholar]
- 10.Washington KT, Craig KW, Parker Oliver D, et al. Family caregivers’ perspectives on communication with cancer care providers. J Psychosoc Oncol. 2019;37(6):777–790. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Kedia SK, Collins A, Dillon PJ, Akkus C, Ward KD, Jackson BM. Psychosocial interventions for informal caregivers of lung cancer patients: A systematic review. Psychooncology. 2020;29(2):251–262. [DOI] [PubMed] [Google Scholar]
- 12.Borges EL, Franceschini J, Costa LH, Fernandes AL, Jamnik S, Santoro IL. Family caregiver burden: the burden of caring for lung cancer patients according to the cancer stage and patient quality of life. J Bras Pneumol. 2017;43(1):18–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Ellis J The impact of lung cancer on patients and carers. Chron Respir Dis. 2012;9(1):39–47. [DOI] [PubMed] [Google Scholar]
- 14.Aubin M, Vézina L, Verreault R, et al. Effectiveness of an intervention to improve supportive care for family caregivers of patients with lung cancer: study protocol for a randomized controlled trial. Trials. 2017;18(1):304. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Ryan PJ, Howell V, Jones J, Hardy EJ. Lung cancer, caring for the caregivers. A qualitative study of providing pro-active social support targeted to the carers of patients with lung cancer. Palliat Med. 2008;22(3):233–238. [DOI] [PubMed] [Google Scholar]
- 16.Schook RM, Linssen C, Schramel FM, et al. Why do patients and caregivers seek answers from the Internet and online lung specialists? A qualitative study. J Med Internet Res. 2014;16(2):e37. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Johnson LA, Blew A, Schreier AM. Health Disparities in Hospice Utilization and Length of Stay in a Diverse Population With Lung Cancer. Am J Hosp Palliat Care. 2019;36(6):513–518. [DOI] [PubMed] [Google Scholar]
- 18.McCarthy EP, Burns RB, Davis RB, Phillips RS. Barriers to hospice care among older patients dying with lung and colorectal cancer. J Clin Oncol. 2003;21(4):728–735. [DOI] [PubMed] [Google Scholar]
- 19.Haley WE, LaMonde LA, Han B, Narramore S, Schonwetter R. Family caregiving in hospice: effects on psychological and health functioning among spousal caregivers of hospice patients with lung cancer or dementia. Hosp J. 2001;15(4):1–18. [DOI] [PubMed] [Google Scholar]
- 20.Demiris G, Oliver DP, Wittenberg-Lyles E. Assessing caregivers for team interventions (ACT): a new paradigm for comprehensive hospice quality care. Am J Hosp Palliat Care. 2009;26(2):128–134. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Parker Oliver D, Albright D, Washington K, Gage A, Mooney M. Hospice caregiver depression: The evidence surrounding the greatest pain of all. J Soc Work End Life Palliat Care. 2013;9(4):256–271. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Demiris G, Ganz F, Han C, Pike K, Oliver D, Washington K. Preliminary testing of the CAregiver Centered Communication Questionnaire (CCCQ). J Pall Care. 2020;35(3):154–160. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Demiris G, DeKeyser Ganz F, Han CJ, Pike K, Parker Oliver D, Washington K. Design and Preliminary Testing of the Caregiver-Centered Communication Questionnaire (CCCQ). J Palliat Care. 2019:825859719887239. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Demiris G, Pitzer K, Washington K, Parker Oliver D. Reliability and validity testing of the Caregiver Centered Communication Questionnaire. J of Nurs Measurement. In Review. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Lowe B, Decker O, Muller S, et al. Validation and standardization of the Generalized Anxiety Disorder Screener (GAD-7) in the general population. Med Care. 2008;46(3):266–274. [DOI] [PubMed] [Google Scholar]
- 26.Kroenke K, Spitzer RL. The PHQ-9: A new depression diagnostic and severity measure. Psychiat Ann. 2002;32(9):509–515. [Google Scholar]
- 27.Oliver DP, Kruse R, Smith J, Washington K, Demiris G. Assessing the Reliability and Validity of a Brief Measure of Caregiver Quality of Life (CQLI-R). J Pain Symptom Manage. 2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Bachner YG, O’Rourke N. Reliability generalization of responses by care providers to the Zarit Burden Interview. Aging Ment Health. 2007;11(6):678–685. [DOI] [PubMed] [Google Scholar]
- 29.Chang VT, Hwang SS, Feuerman M. Validation of the Edmonton Symptom Assessment Scale. Cancer. 2000;88(9):2164–2171. [DOI] [PubMed] [Google Scholar]
- 30.Ferrell B, Rhiner M, Rivera LM. Development and evaluation of the Family Pain Questionnaire. Journal of Psychosocial Oncology. 1993;10(4):21–35. [Google Scholar]