Patient-Caregiver Dyads & End-of-Life Care: Caregiver Personality Disrupts Gender-based Norms

Abstract

Context:

Gender and personality may individually impact end-of-life care. Men often receive more aggressive treatments than women near death, and personality – particularly openness – may be associated with increased care utilization when it diverges from traditional treatment norms. However, research has not examined the interaction of these variables in a dyadic context.

Objectives:

This study examined the dyadic effects of patient gender and caregiver personality on end-of-life care.

Methods:

Using data from the VOICE randomized clinical trial, the present sample consisted of patient-caregiver dyads receiving outpatient care for advanced cancer in Sacramento and Western New York. Analyses assessed whether caregiver personality was associated with gender-based differences in patient chemotherapy or emergency department/inpatient visits in the 30 days before death. Logistic regression examined the interaction between caregivers’ Big Five personality dimensions and patient gender while accounting for patient and caregiver demographic and health characteristics.

Results:

Of a total of 134 patient-caregiver dyads, 19.4% (n=26) of patients received chemotherapy and 47.8% (n=64) had an emergency department/inpatient visit in the 30 days before death. Results demonstrated a significant interaction between caregiver openness and patient gender on receipt of chemotherapy (odds ratio=0.07, p=.006). When caregivers were less open, men were more likely to receive chemotherapy near death, whereas when caregivers were more open, women were more likely to receive chemotherapy near death.

Conclusion:

Results suggest caregiver personality characteristics, particularly openness, might disrupt gender-based treatment norms at end-of-life. Findings demonstrate that patient and caregiver factors can interact to explain patient healthcare utilization.

Introduction

Although most U.S. adults would like to prioritize comfort near the end of life, many instead receive burdensome treatments that undermine their quality of life.(1) More research is needed to clarify the determinants of these “aggressive” end-of-life treatments (i.e., intensive treatments meant to attack, subdue, or remove disease, such as chemotherapy near death), particularly when inconsistent with patient preferences. This study aims to help fill this gap by examining the effects of two possible determinants that have been shown to individually influence healthcare utilization and decision making, gender and personality.(2, 3)

Men in Western cultures tend to receive more aggressive end-of-life treatments than women, consistent with norms regarding masculinity, strength, and toughness.(3) Yet studies assessing the influence of patient gender on end-of-life care have largely overlooked the relational context in which end-of-life decisions tend to occur. Specifically, a growing body of work finds that family caregivers participate meaningfully - both directly and indirectly - in treatment decision-making.(4, 5) Therefore, considering both patient-level and caregiver-level characteristics is necessary for our understanding of end-of-life care. Extensive research demonstrates that patients’ personality contributes substantially to decision-making and healthcare use,(2, 6) but we are aware of no prior research on the implications of caregiver personality for patient treatment utilization at the end-of-life.

Prior research has shown that the relatively stable personality dimension of openness (i.e., open-mindedness, preference for variety, intellectual curiosity, and cognitive flexibility) is associated with treatment utilization, with the direction of effects varying based on whether that utilization goes against traditional treatment norms.(2, 6, 7) Higher levels of openness have been associated with less utilization in terms of routine care visits, emergency department visits, and days spent in skilled nursing facilities.(6,8) However, utilization increases with openness if the care is considered more unconventional. For example, more open individuals are more likely to utilize complementary and alternative treatments,(2) and when residents of U.S. states are more open, they have better palliative care access, a form of supportive care that is typically underutilized and avoided.(7) Openness can disrupt normative behavior.(9) Accordingly, caregiver openness may alter gender-based norms in the context of end-of-life care (i.e., men receiving more aggressive, burdensome treatments while women receive more comfort-focused care).

In these hypothesis-generating analyses, we examine whether the interaction between patient gender and caregiver personality is associated with key indicators of end-of-life care utilization in the Values and Options in Cancer Care (VOICE) study.(1, 10)

Methods

Participants and Procedure

We used data from the VOICE randomized clinical trial, which examined the effects of an intervention on patient-centered communication at two sites in Sacramento and Western New York. The study has been described in full elsewhere (ClinicalTrials.gov identifier: NCT01485627).(10, 11) The sample consisted of patient-caregiver dyads receiving care in outpatient oncology settings, including community based clinics, community hospitals, and academic medical centers.(11) Patients were diagnosed with either a stage IV nonhematologic cancer or a stage III nonhematologic cancer where their oncologist “would not be surprised” if they died in the next 12 months.(10, 11, 12) Patients identified caregivers for participation as a “family member, partner, friend or other individual involved with your health care issues, preferably someone who comes to physician appointments with you”.(11) If more than one caregiver was identified, the primary caregiver was contacted.(10) All participants were at least 21 years of age and provided informed consent. The analytic sample focused on decedents’ end-of-life care. Institutional Review Boards at each study site approved this research (University of Rochester IRB #RSRB0035388).

Measures

Demographic and health characteristics.

Patients and caregivers reported demographic characteristics at study entry. They reported their gender as male or female in response to the item, “What is your gender?”. Patients were only offered binary identities, as data collection began in 2011, and this was culturally normative based on NIH reporting at the time. Patient health information was abstracted from medical records.

Caregiver personality.

At baseline, caregivers completed the Big Five Inventory,(13) which measures self-reported openness, conscientiousness, extraversion, agreeableness, and neuroticism. Total scores were the average of each item (rated 1-5), with higher scores indicating greater levels of the personality dimension.

End-of-life treatments.

We examined receipt of intravenous or oral chemotherapy and presence of any emergency department/inpatient visits in the 30 days before death, key indicators of the aggressiveness of end-of-life treatment.(1) Both variables were abstracted from medical records and coded dichotomously.

Statistical Analyses

We conducted logistic regression analyses in which independent variables were patient gender, caregiver personality (the five dimensions), and their interaction terms. The dependent variables – examined in separate models – were presence of chemotherapy use or emergency department/inpatient visits in the 30 days before death. Covariates were patient and caregiver age and education level, presence of aggressive cancer, caregiver gender, study site, and study arm. Aggressive cancers were predetermined within the VOICE trial based on diagnoses and included lung, gastrointestinal cancers (except colon), and genitourinary cancers (except prostate).(11) Sensitivity analyses accounted for additional demographic and health variables (see supplemental Table 1), and we employed the Benjamini-Hochberg procedure to control for the false discovery rate (Q=.25).

Results

Sample Characteristics

Table 1 presents patient and caregiver descriptive characteristics (N=134 dyads). Dyads received care at medical centers in either western New York (n=97, 72.4%) or Sacramento (n=37, 27.6%). Patients (54.5% female) were primarily White (n=118, 88.1%) and had a mean age of 66.28 years (standard deviation [SD] =10.95). The most common cancer diagnoses were lung (n=35, 26.1%), breast (n=20, 14.9%) and pancreatic (n=20, 14.9%). Half were in each arm of the RCT (n=67, 50.0%).

Table 1.

Patient and Caregiver Characteristics (N=134)

Characteristic	N (%) or Mean (SD)
Patients
Age
<65 years	49 (36.6%)
≥65 years	85 (63.4%)
Gender
Female	73 (54.5%)
Male	61 (45.5%)
Education
High school or less	40 (29.9%)
At least some college	94 (70.1%)
Presence of aggressive cancer
No	57 (42.5%)
Yes	77 (57.5%)
Caregivers
Age
<65 years	70 (52.2%)
≥65 years	64 (47.8%)
Gender
Female	93 (69.4%)
Male	41 (30.6%)
Education
High school or less	35 (26.1%)
At least some college	99 (73.9%)
Personalitya
Openness	3.63 (0.62)
Conscientiousness	4.18 (0.59)
Extraversion	3.26 (0.91)
Agreeableness	4.23 (0.57)
Neuroticism	2.62 (0.75)
End-of-life care received
Chemotherapyb	26 (19.4%)
Emergency Department/Inpatient visitc	64 (47.8%)

Note. SD=standard deviation. Patient-caregiver dyad genders were as follows: 35 (26.1%) woman-woman; 38 (28.4%) woman-man; 3 (2.2%) man-man; and 58 (43.3%) man-woman.

^aPresented as the mean score of 5 items each. Possible scores may range from 1-5.

^bCalculated from a total of N=116 (n=18 missing).

^cCalculated from a total of N=123 (n=11 missing).

Caregivers (69.4% female) were primarily White (n=115, 85.8%) and had a mean age of 60.60 (SD=13.48). They were most commonly patients’ spouses (n=83, 61.9%), children (n=17, 12.7% daughters; n=3, 2.2% sons), or siblings (n=11, 8.2%). Caregivers most frequently reported they knew the patients “extremely well” (a rating of 10 on a 0-10 numeric rating scale; n=68, 50.7%).

End-of-Life Treatments

In total, 19.4% (n=26) of patients received chemotherapy and 47.8% (n=64) had an emergency department visit or inpatient hospitalization in the 30 days before death.

Logistic regression revealed a significant interaction between patient gender and caregiver openness on receipt of chemotherapy (odds ratio=0.07, p=.006; Supplemental Table 1). As shown in Figure 1, when caregivers had low to moderate levels of openness, men were more likely than women to receive chemotherapy in the 30 days before death. However, when caregivers had high levels of openness, men were less likely than women to receive chemotherapy. There were no significant predictors of emergency department/inpatient visits. The chemotherapy finding remained significant after controlling for the false discovery rate.

Figure 1. Caregiver Personality Modifies Gender-based Differences in Patient Receipt of Chemotherapy in the 30 Days Before Death.

When caregivers were more open, male patients were less likely to receive chemotherapy, but female patients were more likely to receive chemotherapy. An interaction effect between patient gender and caregiver openness was tested in a logistic regression model for receipt of chemotherapy using a continuous measure of openness (p=.006; see Supplemental Materials). Caregiver gender did not significantly influence results. Here, openness is trichotomized for ease of graphic depiction (Low Openness [≤3.3]: n=39 caregivers, 21 of male and 18 of female patients; Middle Openness [3.4-3.8]: n=37 caregivers, 14 of male and 23 of female patients; High Openness [≥3.9]: n=40 caregivers, 17 of male and 23 of female patients).

Discussion

This research illustrates that patient gender and caregiver personality interact to explain end-of-life care utilization in advanced cancer. Most prior studies of end-of-life care utilization have focused exclusively on patient demographic characteristics such as gender, with men utilizing more aggressive treatments than women.(3) The present study further contextualizes this relationship. Gender differences in end-of-life care utilization are not necessarily straightforward and may be moderated by caregiver personality, particularly when treatments are relatively discretionary, as is often the case with chemotherapy in advanced cancer. Notably, the results of this study align with traditional Westernized gender-specific socialization: low to moderate levels of caregiver openness, or stricter adherence to conventionality, motivates more aggressive treatment in men and less aggressive treatment in women, in line with what may be seen as more “traditional” gender norms.(3) Conversely, when caregivers are more open, women get more aggressive treatment than men. Ultimately, these findings suggest caregiver personality may promote or discourage adherence to gender stereotypes in patients’ care.

This study is among the first to demonstrate that dyadic interaction effects can explain end-of-life care utilization. To the best of our knowledge, only one other study has documented an interaction between patient and caregiver factors in end-of-life care utilization, showing caregiver treatment preferences more closely predicted end-of-life care intensity in the context of patient cognitive impairment.(14) Our findings suggest that gender and personality are broadly relevant to patient-caregiver dyads, even in situations where cognition does not limit patient capacity for engaging treatment decisions.

Study Limitations

This study is limited by a predominantly White, educated sample of participants, the most recent data points being from 2017, and data infrastructure lacking non-binary gender response options and information about patient personality. The observational data also restricts the ability to make strong causal claims. Additionally, as the study did not explicitly assess attitudes towards gender stereotypes, other factors could also explain the association between openness, gender, and end-of-life treatments. Lastly, the findings exist within the context of a randomized clinical trial, and the results therefore may not be generalizable to individuals who are less likely to participate in trials or receive care at less research-focused institutions.

Conclusions

This study is among the first to observe that patient and caregiver factors can interact to explain end-of-life treatment utilization. These findings reinforce growing clinical endorsement of patient- and family-centered care in serious illness and point to the need for deeper understanding of the influence of the caregiver-patient relationship on medical decision-making. More research is needed to explore how caregiver and patient factors may interact to influence patient care and to inform future interventions promoting preference-concordant decision-making among patients and their families before and at end-of-life.

Key Message:

This article examines the interaction between patient gender and caregiver personality on end-of-life care. Findings indicate caregiver characteristics may influence patient care: when caregivers are more open, men are less likely and women are more likely to receive chemotherapy near death, reversing gender-based treatment norms.