Knowing How to Ask Good Questions: Comparing Latinos and Non-Latino Whites Enrolled in a Cardiovascular Disease Prevention Study

Abstract

Introduction

Latinos face unique challenges engaging with their health care providers for risk management of cardiovascular disease (CVD).

Objective

To better understand differences in how Latinos and non-Latino whites (NLWs) experience CVD care.

Methods

We examined self-reported activation, engagement, confidence, and communication comparing Latinos (n = 194) and NLWs (n = 208). Data were taken from baseline survey assessments of participants in the CREATE Wellness Study (NCT02302612), designed to help patients with poorly controlled CVD risk factors more actively engage in their care. The groups were compared using χ² tests and separate logistic regression models adjusting for age, age and income, and age and educational attainment.

Results

Latinos in this cohort were younger, were less educated, and had lower incomes than did NLWs. In age-adjusted models, Latinos were significantly less likely to report knowing how to ask good questions about their health (71.1% vs 83.7% for NLW, p < 0.01; adjusted odds ratio = 0.49, 95% confidence interval = 0.29–0.83). Further adjustment by educational attainment or income did not attenuate this association. Latinos were also significantly more likely to report positive experiences and confidence with several measures of chronic illness care (adjusted odds ratio range = 1.57–2.01). Further adjustment by educational attainment eliminated these associations.

Conclusion

We found notable differences between Latinos and NLWs in their experience of health care. These results provide insights into how CVD risk management programs can be tailored for Latinos. Interventions to improve patient activation and engagement for Latinos with CVD should emphasize question-asking skills.

INTRODUCTION

Latinos are the largest minority ethnic group in the US and are predicted to comprise up to 30% of the nation’s population by the Year 2050.¹ This diverse population experiences unique challenges to engaging in health care, even among those with health insurance. For example, cultural barriers between Latinos and physicians may result in poor communication, which can consequently lead to poor response to disease prevention interventions.²

Cardiovascular disease (CVD) is the leading cause of death among Latinos.³ Currently, the health care system increasingly relies on disease management programs for chronic conditions such as CVD. Although such efforts are beneficial for some patients, many others still struggle with the management of their chronic illnesses.⁴ Research suggests that a more successful approach to chronic disease interventions should include systematic, patient-oriented efforts that increase patients’ knowledge, skills, and confidence to manage their condition through proactive participation in their own care.⁵ However, limited data exist about knowledge and confidence among Latinos in CVD care programs.

The CREATE Wellness (Changing Results: Engage and Activate to Enhance Wellness) Study (www.ClinicalTrials.gov: NCT02302612) is a recently completed clinical trial designed to help patients with poorly controlled CVD risk factors more actively engage in their care. This randomized clinical trial tested a behavioral intervention for Health Plan members with elevated CVD risk factors (high blood pressure, lipids, and/or glycemic levels) and more than 2 years of not meeting care goals. Here, we report the baseline differences between Latinos and Non-Latino whites (NLWs) enrolled in CREATE Wellness. We compared self-reported, validated survey responses related to patient health states, behaviors, activation, and experience to test the hypothesis that Latinos may have unique barriers to engaging in care programs.

METHODS

Setting

The CREATE Wellness Study was implemented from February 2014 to October 2017 at 4 medical facilities in Kaiser Permanente Northern California, a nonprofit integrated care delivery system providing care for more than 3.8 million members throughout Northern California, including more than 700,000 Latinos. The distribution of members’ demographic and socioeconomic factors is diverse and similar to that of the area population.⁶

Eligibility and Recruitment

Details of the CREATE Wellness clinical trial design have been published previously.⁷ Patient eligibility criteria included membership in a Kaiser Permanente Health Plan, more than 2 years of not meeting care goals for CVD risk management, and ability to provide informed consent in English. Eligible patients received a letter describing the study and a subsequent phone call to schedule an in-person recruitment visit. At this visit, patients independently completed the baseline survey on a desktop computer with a research assistant available to assist as needed.

The Kaiser Permanente institutional review board approved this study. Informed, written consent was obtained from all patients included in the study. All procedures followed were in accordance with the ethical standards of the institutional review board and the Helsinki Declaration of 1975, as revised in 2000.

Survey Instruments

The baseline survey examined the following domains: Patient confidence, activation, engagement, and satisfaction. Participants’ confidence in their ability to communicate with physicians was measured using the Perceived Efficacy in Patient-Physician Interactions (PEPPI) Questionnaire.⁸ The extent of participants’ activation in their own care was measured using the validated Patient Activation Measure Short Form (PAM-13) instrument.⁹ The Effective Consumer Scale (EC-17) explored the attainment of skills, attitudes, and knowledge relating to the self-management of participants’ conditions.¹⁰ The Patient Assessment of Chronic Illness Care (PACIC) instrument was used to measure specific actions or qualities of care that participants reported having experienced in the delivery system.¹¹ In addition, participants answered questions about educational attainment, marital status, yearly household income, and occupational status. All surveys were in English.

Statistical Analysis

Responses to baseline survey items were dichotomized. The 2 patient groups were compared using χ² tests or t-tests as appropriate. We repeated our comparisons using a logistic regression model adjusting for age to account for significant differences between groups. To investigate the potential mediating role of income and educational attainment, we constructed 2 additional models separately adding each of these variables to the age-adjusted model. Survey outcomes that were significant in the age-adjusted model but no longer significant with the addition of income or educational attainment were considered mediated by these factors.

RESULTS

Patient Socioeconomic Status

Survey respondents consisted of 194 Latinos and 208 NLW (Table 1). Latinos in this cohort were significantly younger than NLW (56.7 ± 10.0 years vs 62.6 ± 8.1 years, p < 0.01), had less educational attainment (48.6% vs 74.3% completing college, p < 0.01), had lower incomes (61.5% vs 38.5% with annual income ≤ $50,000/y, p < 0.01), and were more likely to be currently employed (62.4% vs 45.6%, p < 0.01).

Table 1.

Participants’ demographic characteristics (N = 402)

Variable	Latinos (n = 194)	NLW (n = 208)	p value
Age, y (SD)	56.7 (10.0)	62.6 (8.1)	0.01
Women, %	61.9	57.2	0.3
Completed college, %	48.6	74.3	0.01
Married, %	57.7	68.3	0.03
Income ≥ $50,000/y, %	61.5	38.5	0.01
Currently employed, %	62.4	45.6	0.01

NLW = non-Latino white; SD = standard deviation.

Patient Knowledge and Confidence

Despite demographic differences, we found no significant differences between Latinos and NLW in their confidence in making health care decisions, figuring out solutions to new problems, or explaining their main concern to their physicians (Table 2). However, Latinos were significantly less likely to report knowing how to “ask good questions about their health or disease” (71.1% vs 83.7% for NLW, age-adjusted odds ratio [age-aOR] = 0.49, 95% confidence interval [CI] = 0.29–0.83, p < 0.01). In contrast, Latinos were more likely to report that they were confident in helping to prevent or reduce problems associated with their health (96.4% vs 88.0% for NLW, age-aOR = 4.27, 95% CI = 1.72–10.63, p < 0.01).

Table 2.

Survey results comparing Latinos and non-Latino whites (n = 402)^a

Survey domains and questions	Latinos (n = 194)	NLW (n = 208)	p value	Age-aOR (95% CI)
Knowledge and confidence
I know how to ask good questions about my health and my disease (Usually/Always), %	71.1	83.7	0.01	0.49 (0.29–0.83)
I am confident I can help prevent or reduce problems associated with my health (Agree/Strongly agree), %	96.4	88.0	0.01	4.27 (1.72–10.63)
I feel confident explaining my chief health concern to my doctor (Quite confident/Very confident), %	79.4	86.1	0.07	0.60 (0.35–1.04)
Physician-participant interactions
I was asked how my chronic condition affects my life (Most of the time/Always), %	50.0	38.9	0.03	1.57 (1.03–2.38)
I was told how my visits with other types of doctors, like an eye doctor or other specialist, helped my treatment (Most of the time/Always), %	49.0	33.6	0.01	1.80 (1.18–2.75)
I was given a written list of things I should do to improve my health (Most of the time/Always), %	59.8	43.3	0.01	2.01 (1.33–3.06)
I was helped to set specific goals to improve my eating or exercise (Most of the time/Always), %	56.2	42.3	0.01	1.81 (1.20–2.75)
I was helped to plan ahead so I could take care of my condition even in hard times (Most of the time/Always), %	50.5	40.9	0.03	1.60 (1.06–2.44)
I was encouraged to go to a specific group or class to help me cope with my chronic condition (Most of the time/Always), %	47.9	35.1	0.02	1.66 (1.09–2.52)
I was encouraged to attend programs in the community that could help me (Most of the time/Always), %	30.9	21.6	0.01	1.81 (1.13–2.91)

^aResponses were dichotomized (values chosen for analysis shown in parentheses).

aOR = adjusted odds ratio; CI = confidence interval; NLW = non-Latino whites.

Patient Assessment of Chronic Illness Care

Latinos were significantly more likely to report positive experiences with care of chronic illnesses compared with NLWs (Table 2). For example, Latino participants were more likely to report that their providers encouraged them to go to a specific group or class to help them cope with their conditions (47.9% vs 35.1% for NLW, age-aOR = 1.65, 95% CI = 1.09–2.52, p = 0.02), gave them a written list of things they should do to improve their health (59.8% vs 43.3% for NLW, age-aOR = 2.01, 95% CI = 1.33–3.06, p < 0.01), and helped them plan ahead so they could take care of their conditions even in hard times (50.5% vs 40.9% for NLW, age-aOR = 1.61, 95% CI = 1.06–2.44, p = 0.03).

Education and Income as Possible Mediators

We examined the mediating effect of educational attainment and income by adding each variable to our age-adjusted models. Neither variable substantively changed the association of Latino race/ethnicity with lower confidence in posing questions to their physician. In contrast, adding educational attainment (but not income) to our age-adjusted model eliminated or attenuated 4 of 7 significant positive associations between being Latino and satisfaction with physician-patient interactions (Table 3).

Table 3.

Comparison of Latinos vs non-Latino whites in 3 separate models adjusting for age, age and educational attainment, and age and income (N = 402)^a

Survey domains and questions	Age-aOR (95% CI)	Age-and education-aOR (95% CI)	Age-and income-aOR (95% CI)
Knowledge and confidence
I know how to ask good questions about my health and my disease (Usually/Always), %	0.49 (0.29–0.83)	0.57 (0.34–0.97)	0.53 (0.32–0.89)
I am confident I can help prevent or reduce problems associated with my health (Agree/Strongly agree), %	4.27 (1.72–10.63)	4.92 (1.94–12.49)	4.75 (1.86–12.16)
I feel confident explaining my chief health concern to my doctor (Quite confident/Very confident), %	0.60 (0.35–1.04)	0.61 (0.34–1.07)	0.65 (0.37–1.14)
Physician-participant interactions
I was asked how my chronic condition affects my life (Most of the time/Always), %	1.57 (1.03–2.38)	1.37 (0.89–2.11)	1.56 (1.02–2.39)
I was told how my visits with other types of doctors, like an eye doctor or other specialist, helped my treatment (Most of the time/Always), %	1.80 (1.18–2.75)	1.6 (1.04–2.5)	1.83 (1.19–2.82)
I was given a written list of things I should do to improve my health (Most of the time/Always), %	2.01 (1.33–3.06)	1.72 (1.11–2.66)	2.12 (1.38–3.27)
I was helped to set specific goals to improve my eating or exercise (Most of the time/Always), %	1.81 (1.20–2.75)	1.45 (0.94–2.24)	1.79 (1.17–2.73)
I was helped to plan ahead so I could take care of my condition even in hard times (Most of the time/Always), %	1.60 (1.06–2.44)	1.40 (0.91–2.17)	1.63 (1.07–2.51)
I was encouraged to go to a specific group or class to help me cope with my chronic condition (Most of the time/Always), %	1.66 (1.09–2.52)	1.40 (0.90–2.17)	1.75 (1.13–2.69)
I was encouraged to attend programs in the community that could help me (Most of the time/Always), %	1.81 (1.13–2.91)	1.65 (1.01–2.71)	1.85 (1.14–3.00)

^aVariables were dichotomized and values chosen for analysis shown in parentheses.

aOR = adjusted odds ratio; CI = confidence interval.

DISCUSSION

Disease management programs to reduce CVD risk are of high priority for the growing population of Latinos in the US. In our study of patients not meeting all goals of CVD care (blood pressure, lipid, and glycemic control) during the preceding 2 years, we found that Latinos were less confident in knowing how to ask good questions about their health compared with NLW, yet they generally reported more positive interactions with their care team. Lack of confidence in asking questions remained significant even after accounting for the potential mediating effect of educational attainment or income level—traditional measures of socioeconomic status.

These findings can be interpreted within a framework that considers the normative values of Latino culture.¹² The ideals of simpatia (kindness), personalismo (friendliness), and respeto (respect) are important concepts in Latino culture that affect interpersonal relationships and perceptions¹³ and should thus be recognized and built on in clinical settings.¹⁴ Our results specifically underscore how the concept of respeto, defined as the importance of respecting and deferring to those in a position of authority, can influence efforts to help Latino patients engage with their health care team, even after accounting for educational attainment and income level. Prior qualitative research by our group supports these findings, with patients explaining deference to physician authority and expertise as “it’s the doctor’s job to know what to ask.”¹⁵

Prior research findings have demonstrated the impact of cultural differences between Latinos and NLW on 3 core domains of health care that are relevant to CVD risk reduction. For example, Latinos have been shown to have lower adherence to medications for chronic conditions owing to culturally mediated concerns about treatment^16–18 and adverse effects. Similarly, researchers have found that behavioral change strategies related to diet and exercise are more effective when tailored to the needs and beliefs of the Latino population.^19,20 Other work has focused on modifying educational interventions for both Spanish-and English-speaking Latinos.^21–24 Fundamental to all these efforts to influence how Latino patients engage in care is the necessity of effective communication between physicians and patients, which has been shown to be challenging for Latinos in the US.²⁵

A key component of physician-patient communication is the ability of patients to articulate concerns, reservations, and lack of understanding through questions.²⁶ To date, teaching patients the skill of asking questions in the context of medical care has been studied only in a small pilot.^27,28 This work demonstrated that teaching the question-asking skill improved patient activation, which is thought to be in the causal pathway toward better communication and ultimately more effective health care.^29,30 Reluctance to ask questions, especially for patients not meeting CVD management goals, provides a critical and potentially ameliorable barrier to more effective health care for Latinos.

Several limitations of our study should be taken into consideration. Our study cohort consisted of insured patients consenting to participate in a clinical trial of behavioral change, and thus the participants may not be representative of the overall population. However, the finding that question-asking remains an important barrier even among Latinos in this more motivated group suggests that the problem could be even worse among the uninsured and those less willing to participate in research. Another limitation of our study is that all participants spoke English and were willing to answer questions on a desktop computer (with the help of a research assistant), which reduces the generalizability of our findings to Spanish monolingual or low-English proficiency Latinos, who may have additional challenges. Nonetheless, the rich data from our panel of validated survey instruments provide new insights into the needs of Latino patients who are otherwise motivated to improve their health. Finally, social desirability bias may have influenced how participants responded to the survey questions.

CONCLUSION

Our results provide a potential target for future interventions tailored specifically to Latino patients. Although our findings reflect the complexity of developing culturally tailored management strategies, they also provide a foundation that can be built on in future interventions. Implementing techniques to increase patients’ confidence in their communication abilities around question-asking may help address some of the challenges that Latinos face because of culturally mediated health disparities.