Barriers to Pain Management: Incongruence in Black Cancer Caregiving Dyads

Abstract

Context:

To effectively manage cancer pain, there is a need to understand how caregiving dyads appraise symptoms. Dyadic appraisal of symptoms influences whether the dyad perceives the patient’s pain is managed well and whether they are on the same page with their appraisal. Beliefs can act as barriers to the dyadic appraisal.

Objectives:

This secondary data analysis examined incongruence within Black cancer caregiving dyads regarding beliefs about pain management and potential medication side effects using the Barriers Questionnaire-13. Associated factors were also examined.

Methods:

Guided by the Theory of Dyadic Illness Management, dyadic multilevel modeling was conducted with data from 60 Black cancer caregiving dyads to determine the dyadic appraisal of beliefs about pain management and potential medication side effects, which includes the average perception of barriers within the dyad (i.e., dyadic average) and the dyadic incongruence (i.e., gap between patient and caregiver).

Results:

On average, Black cancer caregiving dyads reported moderate barriers regarding pain management (2.262 (SE=0.102, p<.001) and medication side effects (2.223 (SE=0.144, p<.001). There was significant variability across dyads regarding barriers to pain management and medication side effects. Lower patient education and higher patient-reported pain interference were significantly associated with more perceived barriers to pain management and potential medication side effects. Incongruence within dyads regarding barriers to pain management and medication side effects were significantly associated with the caregiver’s report of patient’s pain interference.

Conclusion:

Findings suggest the importance of appraisal that includes both members of Black cancer caregiving dyads regarding pain management.

Introduction

Pain management is one of the predominant areas of cancer caregiving. The process of pain management can involve communicating about pain, decision making, and supporting the cancer patient.¹ Often in cancer pain management, there is someone who assists with pain management such as an informal caregiver, usually a family member or friend.¹ Since most cancer care has shifted from the hospital to an outpatient setting, the number of cancer informal caregivers continues to grow, increasing the extent to which family members are involved in pain management. Most family caregivers report moderate to intense involvement in cancer pain management, including pain appraisal and management of medications.¹

The actions required to manage pain in the home settings are complex, especially if persons are taking opioids to treat their pain. Opioids can be effective for continuous pain and the National Comprehensive Cancer Network endorses maintenance opioid therapy as standard of care, including regularly scheduled and supplemental pain medication doses for breakthrough pain.² While opioids are the recommended treatment for cancer pain management, caregivers of cancer patients may feel ill-prepared in their appraisal of pain and to oversee pain medication management due to their uncertainty with their appraisal and lack of knowledge about the medication.

The caregiver’s role in pain appraisal is important to the pain management of cancer patients.^{3, 4} Although some studies showed that caregivers assess patients’ pain levels similarly, most show that caregivers overestimate patients’ pain.^5–11 In a longitudinal study that examined symptom incongruence between cancer patients and their family caregivers, family caregivers rated the patient’s pain significantly worse than the patient at baseline, and there was no significant change in this incongruence over time.⁵ This overestimation of pain can result in negative outcomes (i.e., worse quality of life, greater distress, and communication problems) for patients and caregivers alike.^5–11

While understanding patient’s and caregiver’s appraisal of pain is necessary, the theory of Dyadic Illness Management focuses on how both members appraise an illness,¹² which may hold promise in navigating cancer pain management. As a result, shared appraisal or more congruent appraisal between dyad members is central to pain management and optimal health outcomes.¹² The way each member of the dyad appraises cancer pain influences how they each engage in behaviors to manage the pain. This process of pain appraisal and pain management is recursive over time, influencing the health of both members of the dyad.¹² Dyadic pain appraisal and pain management activities are heterogenous processes that are influenced by a number of characteristics (i.e., individual, dyadic, family/social, and cultural).

Beliefs about pain and treatment options can be a barrier to pain medication adherence¹³ and reduce pain relief. Patients and caregivers reported multiple barriers related to pain management.¹⁴ Specifically, family caregivers reported concerns about medication administration, handling breakthrough pain and differentiating pain from other symptoms.³ In addition, patients and caregivers reported concerns about addiction to pain medication, potential side effects from the medication, and pain being an inevitable part of the cancer diagnosis trajectory.¹⁴ In a study that examined barriers to effective pain relief based on pain medication administration practices in patients with advanced cancer, racially diverse patients compared to non-Hispanic White patients were more likely to deviate from the prescribed dose.¹⁵

There is limited research on the role of pain beliefs as barriers to pain management in racially diverse cancer patients, and even less on the cancer caregiving experiences within Black cancer dyads. Three studies examined barriers related to pain management in cancer caregiving dyads using versions of the Barriers Questionnaire-13 (BQ-13) and identified associations with beliefs related to pain management in these dyads.^16–18 None of these studies examined dyadic appraisal using within-dyad analysis. Specifically in the study conducted in the United Sates, there was no significant difference in BQ-13 mean scores between cancer patient and family caregiver reports and no correlation to pain intensity, but there was a significant correlation between their BQ-13 total scores.¹⁸ In addition, racial and ethnic differences were found on the BQ-13 total scores.¹⁸ Based on BQ-13 total scores, patients with a Hispanic heritage reported greater barriers than non-Hispanic patients; caregivers from minority groups (characterized as African American, Asian American, Native American, Other, and Multiracial) reported significantly more barriers than non-minority caregivers.¹⁸ However, when African American patients or caregivers were compared to non-Hispanic White patients or caregivers neither BQ-13 total scores were significantly different.¹⁸ The current study sought to examine the dyadic appraisal among Black cancer patients and their family caregivers beliefs’ of barriers regarding pain management and potential medication side effects and associated factors. Furthermore, this study used dyadic multilevel modeling to 1) describe dyadic averages and incongruence regarding concerns about pain management and potential medication side effects total scores and 2) examine associations of individual-level factors of Black cancer caregiving dyads.

Methods

Participants and Setting

This secondary data analysis used the data from the parent study that explored congruence of symptom assessment between Black cancer caregiving dyads focusing on perceptions of patients’ cognitive, depressive, and pain symptoms. A convenience sampling of 60 Black cancer caregiving dyads were recruited between April 2016 and March 2017. These participants were recruited from medical oncology, radiation oncology, and palliative care clinics of an urban safety-net hospital and from a tertiary cancer center. Eligible patients were: 1) 21 years or older; 2) identified as Black or African American; 3) diagnosed with cancer; 4) prescribed and reported taking opioids for cancer pain (i.e., reported taking at least one dose each day for the past week); 5) mentally competent as reported by their clinical provider; and 6) living in the metro Atlanta area. Patients were excluded if they had surgery within the past month since the post-operative pain experience may be different. Inclusion criteria for caregivers were: 1) 21 years or older; 2) identified as the main caregiver by the patient; 3) living with the patient for the last 6 months; 4) not receiving a salary for caregiving duties; and 5) mentally competent. The Institutional Review Board at Emory University and the Research Oversight Committee at Grady Health approved this study prior to recruitment.

Study Procedures

After receiving informed consent, research staff and the study participants scheduled the data collection visits at a mutually agreed upon time and place (e.g., patient’s or caregiver’s home or a private room at the clinic). Each visit included a one-time in-person session with the patient and caregiver separately. In most visits, the surveys were interviewer administered (i.e., questions read aloud). For visits that participants preferred to answer the questions independently using paper/pencil, the research staff checked the surveys for completeness at the visit. Both the patient and caregiver visits took approximately 30 minutes to complete.

Instruments

Barriers Questionnaire-13.

The Barriers Questionnaire-13 (BQ-13) measures beliefs regarding pain and pain management that may serve as barriers to managing pain.^{18, 19} The BQ-13 includes two subscales—pain management and side effects.²⁰ The pain management subscale consists of eight items, such as “having pain means the disease is getting worse.” The side effects subscale consists of five items, such as “confusion from pain medicine is really a bother.”²⁰ For both subscales, the responses of each item range from 0 (“do not agree”) to 5 (“agree very much”). Higher scores reflect that the responder possesses more concerns about pain management or more concerns about potential pain medication side effects. Each subscale was completed by the cancer caregiving dyads, with each member of the dyad reporting their own beliefs about pain management and potential pain medication side effects independently. The pain management and side effects subscales have been validated and demonstrated fair internal consistency with Cronbach’s alpha of 0.63 and 0.77, respectively.²⁰ In this study, the Cronbach’s alpha of the pain mangement and side effects subscales were 0.63 and 0.70 for patients and 0.69 and 0.82 for caregivers, respectively.

Individual-Level Factors.

Individual-level factors are those that are related to only one member of the dyad, which included demographic and clinical information, assessment of patient’s pain treatment satisfaction, pain severity, and pain intenferene.

Demographic and Clinical Information.

Self-reported demographic data were collected from both members of the Black cancer caregiving dyads. Demographic and clinical information included age, gender, educational level, employment status, and annual individual incomes. Cancer type, presence of metastasis, and comorbidities were gathered by review of the electronic medical records.

Pain Satisfaction.

Using a single item, patients were asked to rate their satisfaction with the results of their overall pain treatment. The pain rating was from 10-point Likert scale with 0=dissatisfied and 10=extremely satisfied.

Brief Pain Inventory.

The Brief Pain Inventory (BPI) was used to measure cancer pain within the past 24 hours. The BPI consists of two subscales: pain severity and pain interference. Pain severity is assessed as “worst”, “on average”, and “current” on a 10-point Likert scale.²¹ These three scores were averaged to determine the pain severity.²² A higher score is interpreted as more severe pain. Pain interference is calculated by averaging seven items that measure the degree to which pain hinders mood, general activities, and relationships. In addition, the BPI has one question about percentage of pain relief by analgesics.²¹ The BPI has been validated for assessing pain in cancer patients²³ and tested in ethnic and racial minority groups.²⁴ A high internal consistency of the BPI subscales has been demonstrated with Cronbach’s alpha of 0.85 for pain severity and 0.87 for pain interference.²² The Cronbach’s alpha of the pain severity and pain interference in our sample were 0.75 and 0.87 for patients and 0.71 and 0.87 for caregivers.

Statistical Analyses

Descriptive statistics were computed for the individual-level factors and dyadic appraisal variables (i.e., dyadic average and dyadic incongruence). The dyadic average is the average perception of barriers reported by the patient and caregiver; the dyadic incongruence is the gap or difference between patient and caregiver. Bivariate analysis was conducted with the individual-level factors (i.e., pain severity, pain interference, age, and educational status from both patients and caregivers, percentage of pain relief and satisfaction with results of pain treatment from the patients only) to determine which variables would be entered in the model as predictors. Multilevel modeling was used to analyze dyadic averages and incongruence and is considered the optimal approach to examining dyadic appraisal.^25–27 This approach has been used in prior studies of individuals with chronic pain and their spouses²⁸ and in caregiving dyads of older adults living with dementia.^{26, 29} In particular, HLM 8 software³⁰ was used to examine dyadic averages and incongruence in concerns about pain management and potential pain medication side effects in two separate models. First, a multilevel within-dyad model was used to estimate dyadic average and dyadic incongruence scores within dyads from individual patient and caregiver raw data.

Using a dyad indicator variable coded −.5 for patients and .5 for caregivers, the intercept represents the average barrier scores within the caregiving dyad and the slope represents the incongruence (gap between the patient and caregiver perceptions) regarding barrier scores. Given the coding of the dyad indicator variable, a negative slope would indicate patients report more concerns than caregivers. Both the intercept and slope parameters are considered fixed effects (i.e., population average estimates). Next, a between-dyad model included predictors to explain variability between or across dyads in dyadic average and dyadic incongruence in the barrier scores. Potential predictors were selected based on the pain literature^{31, 32} and bivariate analysis. HLM uses maximum likelihood estimation to deal with missing data, which are assumed to be missing at random.³³ For a more detailed description of the dyadic incongruence model refer to this article.³⁴

Results

Demographic and Clinical Characteristics

In this study, 60 Black cancer patients were 57±11 years old on average, with almost half identifying as women (46%). Over a quarter of them (28%) were diagnosed with breast cancer, a third (33%) with a comorbid diagnosis of HIV/AIDS, and the majority had greater than a high school education (52%). A majority of patients reported their annual individual incomes was less than or equal to $30,000 (62%). Black caregivers were 51±14 years old on average and predominantly women (60%). Most of the caregivers were non-spouse/partner caregivers (40%) and less than half had a high school education (46%). Table 1 provides additional sample characteristics.

Table 1.

Demographic and Clinical Characteristics of Black Cancer Patients and Their Caregivers

Variables	Patients (N=60)	Caregivers (N=60)
	Mean ± SD or n (%)	Mean ± SD or n (%)
Age in years	56.97 ± 10.77	50.63 ± 13.90
Gender (Women)	36 (60.0)	33 (60.0)a
Education (greater than HS)	31 (51.7)	25 (46.3)a
Individual incomea
< $10,000	13 (35.1)	---
$10,000 to < $30,000	18 (48.6)	---
≥ $30,000	6 (16.2)	---
Relationship to patientsa
Spouse/partner	---	22 (40.0)
Child	---	10 (18.2)
Parent	---	8 (14.5)
Friend	---	4 (7.3)
Otherb	---	11 (20.0)
Currently working	8 (13.3)	23 (41.8)a
Cancer type
Breast	18 (30.0)	---
Multiple myeloma	14 (23.3)	---
Gastrointestinal	7 (11.7)	---
Lung	5 (8.3)	---
Prostate	5 (8.3)	---
Laryngeal	3 (5.0)	---
Otherc	8 (13.3)	---
Presence of metastasisa (Yes)	29 (58.0)	---
Comorbid chronic diseased	32 (53.3)	29 (53.7)a
HIV/AIDS	20 (33.3)	1 (19)
Arthritis	8 (13.3)	7 (13.0)
Diabetes	7 (11.7)	9 (16.7)
Kidney disease	6 (10.0)	2 (3.7)
Hypertension	4 (6.7)	22 (40.7)
Heart disease	3 (5.0)	1 (19)
Respiratory disease	2 (3.3)	4 (7.4)
Brief Pain Inventorya
Pain Severityf	6.61 ± 1.84	6.80 ± 1.75e
Pain Interferenceg	5.62 ± 2.35	6.22 ± 2.39e
Barriers Questionnairea	30.81 ± 12.15	29.12 ± 14.53
Pain Managementh	2.35 ± 0.97	2.18 ± 1.09
Side Effectsi	2.24 ± 1.23	2.20 ± 1.57

Note: HS= high school; SD= standard deviation; HIV= human immunodeficiency virus; AIDS= acquired immunodeficiency syndrome

^aOnly respondents who answered this question were included.

^bOther relationships to patients include brother, sister, girlfriend, sister-in-law, aunt, fiancé, and boyfriend.

^cOther cancer types include leukemia, lymphoma, pancreatic, liver and tongue.

^dRespondents selected all that apply.

^eCaregiver proxy assessment of patient pain.

^fMinimum/maximum scores for patients 2/10 and for caregivers 0.6/10.

^gMinimum/maximum scores for patients 0/10 and for caregivers 0/10.

^hMinimum/maximum scores for patients 0/4.75 and for caregivers 0/4.29.

ⁱMinimum/maximum scores for patients 0/4.80 and for caregivers 0/5.

Mean Patient and Caregiver BQ-13 Scores

As presented in Table 2, mean scores greater than 3 indicate areas of higher concern. For both patients and caregivers, “I do not like taking pills” was the highest concern. In addition, three other areas of higher concern for patients were “constipation from pain medication,” “focus on curing illness than time controlling pain,” and “get addicted to pain medications easily.” At the item level, only one concern “pain medicine often makes you do embarrassing things” was a significantly higher concern for caregivers than patients.

Table 2.

Mean (SD) for Each Barriers Questionnaire 13 Items Reported by Patients and Their Caregivers

Variables	Patients	Caregivers
	Mean ± SD	Mean ± SD	p-value
Pain Management	2.35 ± 0.97	2.18 ± 1.09
Having pain means the disease is getting worse.	2.14 ± 1.77	1.78 ± 1.69	.318
I do not like taking pills.	3.36 ± 1.87	3.04 ± 1.98	.422
Pain medication cannot really control pain.	1.84 ± 1.75	1.55 ± 1.91	.425
People get addicted to pain medication easily.	3.37 ± 1.78	2.98 ± 2.00	.315
It is important to be strong by not talking about pain.	1.55 ± 1.85	1.18 ± 1.75	.297
It is more important for the doctor to focus on curing illness than to put time into controlling pain.	3.25 ± 1.81	2.63 ± 1.98	.074
You take pain medicine when you have some pain, then it might not work as well if the pain becomes worse.	2.74 ± 1.93	2.38 ± 1.97	.334
It is easier to put up with pain than with the side effects that come from pain medicine.	1.35 ± 1.63	1.92 ± 1.95	.121
Side Effects	2.24 ± 1.23	2.20 ± 1.57	.855
Drowsiness from pain medication is really a bother.	2.47 ± 1.67	1.95 ± 1.89	.061
Confusion from pain medication is really a bother.	1.77 ± 1.89	1.80 ± 1.90	.922
Nausea from pain medication is really distressing.	2.52 ± 1.96	2.62 ± 2.07	.740
Pain medicine often makes you say or do embarrassing things.	1.03 ± 1.41	1.70 ± 1.91	.041
Constipation from pain medicine is really upsetting.	3.43 ± 1.90	2.95 ± 2.08	.122

Dyadic Incongruence about Concerns for Pain Management

Results of the fixed effects of the within-dyad model (Table 3) indicated that, on average, caregiving dyads rated their beliefs about pain management as 2.262 (SE=0.102, p< .001), indicating a moderate level of barriers related to pain management reported within dyads (scale from 0–5). The average incongruence score of beliefs about pain management was −0.183, which was not significantly different from zero (Table 3). Random effects showed significant variability around both the dyadic average scores (χ²=150.696, p< .001) and incongruence scores (χ²=125.062, p< .001), indicating that some dyads reported more barriers and others reported less barriers as well as some dyads had high levels of incongruence and some dyads had very little to no incongruence.

Table 3.

Multilevel Model of Incongruence in Barriers to Pain Management

Fixed Effects	Coefficients (SE)
	Within-Dyad Model	p-value	Between-Dyad Model	p-value	Effect size
Dyad average
Intercept	2.262 (0.102)	< .001	2.675 (0.133)	< .001
Patient pain interference			0.122 (0.041)	.004	.42
Patient age			0.012 (0.009)	.202	.19
Patient education			−0.495 (0.188)	.012	.37
Caregiver report of patient pain interference			0.025 (0.040)	.526	.10
Caregiver age			−0.012 (0.006)	.078	.27
Caregiver education			−0.204 (0.187)	.282	.16
Dyad incongruence
Intercept	−0.183 (0.186)	0.329	−0.362 (0.276)	.198
Patient pain interference			−0.091 (0.085)	.289	.16
Patient age			−0.035 (0.019)	.075	.27
Patient education			−0.604 (0.393)	.132	.23
Caregiver report of patient pain interference			0.167 (0.083)	.049	.30
Caregiver age			−0.0001 (0.013)	.995	.00
Caregiver education			0.560 (0.391)	.160	.21

Note: effect size r (correlation coefficient)= $\sqrt{t^2/\left(t^2+df\right)}=.10\left(\text{small}\right)$, .30 (medium, .50 (large).

The between-dyad model included relevant predictors (i.e., patient age, caregiver age, caregiver education). We found that lower levels of patient education (β= −0.495, SE=0.188, p= .012) and higher levels of patient-reported pain interference (β=0.122, SE=0.041, p= .004) were significantly associated with the dyad reporting more barriers related to pain management (Table 3). For dyads with patients who had greater than a high school education, there was a 0.5 decrease in their barriers related to pain management. For incongruence, caregivers’ perception of the patient’s pain interference (β=0.167, SE=0.083, p= .049) was significantly associated with incongruence regarding beliefs related to pain management, controlling for other variables. The higher the caregiver rated the patient’s pain interference, the less incongruent (i.e., more congruent) the beliefs within the dyad.

Dyadic Incongruence about Concerns to Medication Side Effects

Results of the fixed effects of the within-dyad model showed that, on average, caregiving dyads rated beliefs related to potential side effects as 2.223 (SE= 0.144, p< .001), indicating a moderate level of barriers regarding potential pain medication side effects within dyads (scale from 0–5). The average incongruence score for beliefs about potential side effects was −0.040, which was not significantly different from zero (Table 4). Random effects indicated significant variability around the dyadic average scores (χ²=308.109, p< .001), indicating some dyads reported more barriers and some reported less barriers. In addition, there was significant variability around the average dyadic incongruence score (χ²=173.135, p< .001), indicating some dyads experienced high levels of incongruence and some dyads reported low to no incongruence.

Table 4.

Multilevel Model of Incongruence in Barriers to Management of Side Effects

Fixed Effects	Coefficients (SE)
	Within-Dyad Model	p-value	Between-Dyad Model	p-value	Effect size
Dyad average
Intercept	2.223 (0.144)	< .001	2.651 (0.187)	< .001
Patient interference			0.158 (0.058)	.009	.38
Patient age			−0.005 (0.013)	.732	.05
Patient education			−0.593 (0.271)	.034	.34
Caregiver report of patient interference			0.056 (0.056)	.326	.15
Caregiver age			−0.008 (0.009)	.367	.14
Caregiver education			0.222 (0.269)	.413	.12
Dyad incongruence
Intercept	−0.040 (0.217)	.854	−0.073 (0.345)	.834
Patient interference			−0.198 (0.107)	.070	.27
Patient age			−0.018 (0.024)	.464	.11
Patient education			−0.812 (0.500)	.112	.24
Caregiver report of patient interference			0.212 (0.103)	.046	.30
Caregiver age			0.006 (0.017)	.715	.06
Caregiver education			0.950 (0.500)	.062	.28

Note: effect size r (correlation coefficient)= $\sqrt{t^2/\left(t^2+df\right)}=.10\left(\text{small}\right)$, .30 (medium, .50 (large).

After controlling for the selected predictors (patient age, caregiver age, caregiver education) in the between-dyad model, lower levels of patient education (β= −0.593, SE= 0.271, p= .034) and higher levels of patient-reported pain interference (β=0.158 , SE=0.058, p= .009) were significantly associated with dyads reporting more barriers related to potential pain medication side effects. For dyad incongruence, caregivers’ reports of patient pain interference (β= 0.212, SE=0.103, p= .046) were significantly associated with incongruent beliefs about potential side effects, increasing the magnitude of incongruence within the dyad, but also changing the direction of the incongruence. In other words, the higher the caregiver perceived the patient’s pain interference, the larger the gap between their beliefs and the more likely the caregiver was to report more barriers regarding potential side effects compared to the patient.

Discussion

Pain management is an important component of cancer caregiving, which often involves the appraisal from both members of the dyad. Congruent appraisal has been suggested to result in better illness management and health outcomes for both members of the dyad.³⁵ Specifically, understanding (in)congruent beliefs about pain management and potential side effects as barriers to pain management within the dyad may hold promise in effectively managing the patients’ pain. This study is the first to examine the beliefs about pain management and potential medication side effects in Black cancer caregiving dyads considering the dyadic appraisal of pain. We found that 1) on average, dyads reported a moderate level of barriers regarding pain management and potential medication side effects, 2) there was significant variability around the dyadic averages and incongruence scores, 3) factors associated with the dyadic average within dyads about pain management and potential medication side effects were the patient’s educational level and patients’ report of their own pain interference (highlighting the importance of patient factors), and 4) the caregiver’s report of the patient’s pain interference was associated with less incongruence within the dyad about pain management but more incongruence within the dyad about potential medication side effects (highlighting the importance of caregiver factors).

Previous research^{19, 21, 36} supports our findings that pain management and potential side effects of pain medication are of moderate concern. Brooks et al³⁷ interviewed cancer patients and caregivers and found that they were especially concerned about opioids use and the stigma around these drugs. The BQ-13 questionnaire probes at beliefs that providers may not routinely address in their assessment of the patient’s pain experience, and the nature of these beliefs may further complicate educating patients about pain management. Review of the individual BQ-13 items may result in a more tailored approach to pain assessment and potential barriers to pain relief, which may especially benefit Black cancer caregiving dyads since cancer pain has been stigmatized in many Black communities and considered an unavoidable part of the diagnosis.³⁸

In addition, the findings support the importance of not assuming a “one size fits all” approach, highlighting the importance of examining heterogeneity within racial/ethnic minority groups. Wilkie et al¹⁸ found that when comparing non-Hispanic White caregiving dyads to African American dyads, there was no significant differences between the two racial groups’ reports of barriers to pain management. Without further investigation, using within-group analysis our findings would have been missed. Whitfield et al³⁹ supports the importance of within racial and ethnic group examination and not just comparison studies that may miss subtle nuances.

The dyadic averages of barriers regarding pain management and potential medication side effects were associated with dyads whose patients reported lower educational levels and whose patient’s reported higher levels of their own pain interference. Lower educational levels have been significantly associated with lower health literacy,⁴⁰ which could lead to delayed cancer screening and inappropriate decision making with cancer treatment and symptom management. Health care providers must consider health literacy factors as they work to ensure pain management consultations are undertaken in a way to optimize comprehension and learning for both the patient and caregiver. Our findings support the role of both members of Black cancer dyads when understanding barriers related to pain management and potential medication side effects as well as the degree of incongruence within and across these dyads.

In addition to the lower educational levels of the patient, dyads whose patients reported higher levels of pain interference reported significantly more barriers to pain management and potential side effects, highlighting the impact of pain on daily functioning. Higher levels of functional interference caused by pain in the areas of daily activity, mood, walking, sleeping, ability to do work, enjoyment of life, and relationships with others as reported by patients were associated with lower quality of life⁴¹ and can indicate poorly managed pain impacting the patient’s day to day life. The relationship of pain interference and barriers perhaps may indicate that ineffective pain relief increases concerns about pain management and potential side effects. It is of interest to note, pain interference, not pain severity, was significantly associated with concerns about pain management and potential medication side effects in the bivariate analysis. Similarly, Wilkie et al¹⁸ found no association with pain severity and barriers related to pain management. While pain severity is more commonly assessed in the clinical setting, pain interference may be easier to quantify than pain severity making this a more sensitive indicator.

Research suggests there is a need for more cultural training regarding pain assessment by healthcare providers.⁴² Two cultural phenomena of interest are the Strong Black Woman/Superwoman Schema^{43, 44} and John Henyrism.⁴⁵ The Strong Black Woman/Superwoman Schema describes characteristics often reported by Black women regarding self-silencing their own physical and psychological pain through a cultural mandate to exude strength.^{43, 44} The John Henryism hypothesis describes the active coping that many Black men face in their high-stress environments that negatively influence their health.⁴⁶ In our sample of Black cancer caregiving dyads, pain interference was able to capture significant pain experiences despite these potential cultural phenomena highlighting the need to assess pain interference in addition to pain severity. Further work is needed to explore the interaction between culture, pain interference, and beliefs about pain and its treatment.

Our findings demostrate incongruence within dyads about pain topics that have implications for managing pain and highlight the need for providers to include both members of the dyad in education and decisions about pain treatment plans. Incongruence within Black cancer caregiving dyads was significantly associated with the caregiver’s report of the patient’s pain interference. Dyads whose caregivers reported greater pain interference experienced by the patient, controlling for the patient’s report of their pain interference, were more congruent regarding barriers about pain management. Caregivers can witness the effect of pain on physical functioning and want to intervene to provide pain relief. Congruence in dyadic appraisals has been shown to suggest the members of the dyad are “on the same page” and working as a team.¹² As a result, more congruent dyads have been found to report better well-being and self-management within the dyad.¹² However, there was significantly more incongruence regarding barriers related to potential medication side effects in dyads whose caregivers reported greater patients’ pain interference. One potential explanation is that caregivers may be more concerned with the potential side effects than patients who have weighed the cost/benefit of the pain medication. For example, a caregiver may worry that “Drowsiness from pain medication is really a bother” (an item from the BQ-13), but the patient experiencing pain would rather use the medication to decrease the suffering from the pain and live with the drowsiness.

Our findings show the importance of using a multilevel modeling approach when examining cancer caregiving dyads. This approach provides the opportunity to move beyond average scores and stresses the importance of examining heterogeneity in Black caregiving experiences.^{47, 48} For example, in this study the average showed no incongruence. A more traditional approach to the analysis might have concluded that there was no incongruence in Black cancer caregiving dyads, implying they all agree. However, using this multilevel modeling approach to incongruence highlighted the variability across dyads. Some dyads had little to no incongruence and others had much more incongruence. This finding is important for clinicians and providers when having discussions about pain management and for researchers who are looking to develop tailored interventions for this population.

Despite the strengths of our study, there are limitations. Due to our small sample size, we were limited in the number of variables we could include in the models. Future studies with larger samples are needed to allow for more predictors in the models to provide a comprehensive understanding of factors associated with incongruence within dyads regarding these salient issues. In addition, the study is cross-sectional, which does not allow for the determination of causal inferences in the predictors or incongruence associated with the concerns about pain management and potential side effects of pain medications. Lastly, this is a convenience sample so these results may not be generalizable to all Black cancer caregiving dyads.

Our findings regarding incongruence within Black cancer caregiving dyads related to pain management and potential medication side effects highlight the importance of both members of the dyad in navigating pain management. Caregiving is often described as an interdependent or dyadic process,^{26, 49} which is supported by our findings. Thus, the inclusion of both members of Black cancer caregiving dyads—especially caregivers who will assist in managing pain medications—in the patient’s visits may be necessary to gain an understanding of concerns about pain treatment. This understanding may create a more culturally tailored experience for Black cancer caregiving dyads that supports effective pain management of Black cancer patients.