Effective diabetes care, with its multitude of patient self-management tasks, requires effective communication and collaboration between patient and physician. Studies examining the relationship between various aspects of the patient physician interaction and diabetes outcomes have focused primarily on patient-centered outcomes such as satisfaction and overall self-rated health,1 or on reports of self-management behaviors such as medication adherence or diet and exercise.2–5 Fewer studies have examined the relationship between patient physician interactions and clinical outcomes such as glycemic control or blood pressure and lipid control.6–10
Understanding the relationship between the patient physician relationship and diabetes outcomes may be particularly important in the care of the urban poor. Racial and ethnic minorities, people with low socioeconomic status or education, and the uninsured are all more likely than their counterparts to develop diabetes and its related complications.11 Culturally competent care, defined by the National Quality Forum as the provision of care that is “safe, patient and family centered, evidence based, and equitable” for diverse populations12 has been posited to improve patient outcomes, including diabetes care; however, studies exploring this hypothesis are few and evidence for this link is scant.13,14 The paucity of research may be due, in part, to the historical lack of validated measures of cultural competent care with which to examine this link.
We used a recently developed and validated multidimensional measure of culturally competent care to examine the relationships between the 3 core aspects of culturally competent care and 3 diabetes-related clinical outcomes in an ethnically diverse sample of low-income patients. We hypothesized that glycemic, lipid, and blood pressure control, which often require a combination of physician-guided medication use and sustained lifestyle modification, would be sensitive to patient report of physician communication behavior, trust in physician, and communication about health promotion activities.
Study Design, Setting, and Participants
We analyzed data from the Immigration, Culture, and Health Care (ICHC) study. The ICHC study is a cross-sectional study of African American, Spanish-speaking and English-speaking Mexican American, and non-Hispanic white adults with diabetes who received care in 9 free-standing or hospital-based safety-net clinics in the San Francisco Bay Area and Chicago in 2008–2009. The main purpose of the ICHC was to explore factors that impact diabetes self-management and health outcomes in minority populations. To be included in the study, patients had to have type 2 diabetes, be 18 years of age or older, and speak English or Spanish. Exclusion criteria, assessed by trained interviewers before study enrollment, included cognitive impairment, active substance abuse, or psychosis severe enough to interfere with survey administration. Recruitment was stratified by race/ethnicity and patient language to ensure a diverse sample. The participation rate among eligible patients was 91%.
After providing written informed consent in English or Spanish, participants completed an in-person survey with a trained, bilingual, and bicultural research assistant. Clinical data was abstracted from participants’ electronic health record. The values for glycosylated hemoglobin (A1C), low-density lipid (LDL) cholesterol, and systolic blood pressure (SBP) recorded within 1 year prior and closest to the date of the interview were abstracted. The Institutional Review Boards at the University of California San Francisco, the Cook County Health and Hospital System, and the participating institutions in Chicago and the San Francisco Bay Area approved all study activities.
We used 3 domains from the Consumer Assessment of Healthcare Providers and Systems Cultural Competence (CAHPS-CC) as our measures of core aspects of cultural competence. CAHPS-CC is a 26-item set designed to measure patients’ overall experience of their physician’s interpersonal and cultural competence as well as their experience of their physician’s office. Items were developed through extensive use of focus groups, cognitive interviewing, field testing, and rigorous translation described in detail elsewhere.15 We used the 3 CAHPS-CC domains that were psychometrically sound among both English and Spanish speakers in our participant population.16 They represent 3 core aspects of cultural competence that may be particularly germane to diabetes outcomes: Doctor Communication-Positive Behaviors (5 items), Trust (5 items), and Doctor Communication-Health Promotion (4 items). The Doctor Communication-Positive Behaviors domain was comprised of items which asked patients about the extent to which their provider exhibited several aspects of good communication behavior, such as listening carefully, showing respect, and spending adequate time with the patient. The Doctor Communication-Health Promotion domain was comprised of items that asked patients about the extent to which their provider discussed a number of health and wellness behaviors, including physical activity, depression, and healthy diet. The CAHPS-CC Trust domain was comprised of items aimed to assess the patient’s trust in their physician, including the extent to which they feel like they can tell their provider anything, tell their provider the truth about their health, and trust their provider with their medical care. The individual questions making up the 3 domains are presented in Appendix 1.
Items comprising each domain were transformed to a 0 to 100 scale, with a score of 100 representing the most culturally competent response for each item. We derived a domain score by calculating the mean of all items in the domain; therefore, a domain score of 100 indicates no responses on any item in the domain were consistent with less cultural competence. As found in previous studies of patient assessments of communication quality,17 scores for each of the domains were positively skewed. We therefore divided the scores for each domain into quartiles and dichotomized patients’ scores as “High” if scores fell into the top quartile and “Low” for all other scores.
Our primary outcomes were poor glycemic, LDL cholesterol, and SBP control. We dichotomized as “poor control” A1C≥8%; LDL cholesterol>100 mg/dL and SBP≥140 mg. We chose a more liberal (worse) cutoff of A1C and SBP to indicate poor control than those recommended by established guidelines for good control18 in keeping with current trends to compare those at high risk to those at low risk for complications related to poor control. This approach may be more sensitive in capturing populations at highest risk.8,19,20
Other survey variables used as covariates in our analysis included several known or hypothesized patient-level risk factors associated with diabetes control: age (y), sex (male or female), race/ethnicity, language, diabetes duration (years since diagnosis), number of comorbidities, and depression. Patients reported their ethnic identification as Mexican/Mexican American, white, or African American. Patients who specified that English was their primary language or that they spoke English “well” or “very well,” were categorized as “English speaking,” whereas patients who specified that English was not their primary language and that they spoke English “not well” or “not at all” were considered “Spanish speaking.” Patient report of comorbid illnesses (coronary artery disease, stroke, cancer, hypertension, arthritis, and hyperlipidemia) was used to generate a continuous variable of total number of comorbidities (0–6). Depression was assessed with the Patient Health Questionnaire-9 (PHQ-9), a 9-item instrument validated both in English and Spanish21 and used to identify depressive symptoms. In keeping with common practice, a PHQ-9 score of ≥10 was considered indicative of depression.22
For each of the primary clinical outcomes we used χ2 tests to evaluate bivariate associations between poor clinical control and each of the 3 CAPHS-CC dichotomized domains. We used multivariate logistic regression models that included risk factors potentially associated with diabetes clinical outcomes (age, sex, race/ethnicity, language, diabetes duration, comorbidities, and PHQ-9 depression score) to isolate the impact of reports of each domain of culturally competent care on poor clinical control. From these logistic regression analyses, we reported the odds of poor clinical control for participants reporting a High vs. Low score on each domain of culturally competent care. All analyses were performed using SAS 9.2.
We repeated all analysis using values of A1C≥7% and SBP≥130 mg as measures of poor control to assess sensitivity associated with our definition of poor control.
The ICHC recruited 711 patients. To standardize the exposure, only patients who reported having a regular primary care physician were included in this analysis. This yielded a racially and ethnically diverse group of 600 participants with type 2 diabetes that were predominantly Mexican American (53%), English speaking (75%), low income, and with low educational attainment (73% having a high school/general educational development or less). Patients were evenly distributed by sex and recruitment city. The rate of depression was high; 35% met criteria for current depression using the PHQ-9. Poor clinical outcomes were common, with 47% of the sample having poor glycemic control (based on 558 patients with A1C available), 36% with poor lipid control (based on 480 patients with LDL measurements), and 33% having poor SBP control (based on 595 patients) (Table 1).
In bivariate analyses (Table 2), A1C was significantly related to the domains of Trust and Doctor Communication-Health Promotion but not Doctor Communiation-Positive Behaviors. Patients who reported higher Trust were less likely to have poor control than those who reported low Trust (41.2% vs. 53%, P=0.005). In contrast, Doctor Communication-Health Promotion was inversely associated with poor glycemic control; patients who reported high scores on Doctor Communication-Health Promotion were more likely to have poor glycemic control (54.3% vs. 44%, P=0.03). SBP was significantly related to the domain of Doctor Communication-Positive Behaviors; patients reporting high Doctor Communication-Positive Behaviors were more likely to have poor SBP control (39.7% vs. 29.1%, P=0.007). There was no relationship between any of the 3 domains and poor LDL cholesterol control.
Table 3 displays the adjusted odds of poor clinical control by domain of cultural competent care. After adjustment for sociodemographic and clinical factors, Trust remained inversely associated with poor glycemic control [odds ratio (OR), 0.59; 95% confidence interval (CI), 0.41–0.84] and Doctor Communication-Health Promotion remained directly related to poor glycemic control (OR, 1.49; 95% CI, 1.02–2.19), indicating that patients who reported higher trust in their physician were less likely to have poor glycemic control, whereas those reporting high health promotion communication were more likely to have poor control (Table 3). There continued to be no relationship between any of the 3 domains of culturally competent care and SBP control or LDL cholesterol control after adjustment for sociodemographic and clinical factors.
Reanalysis using lower (tighter) cutoffs for glucose showed similar point estimates, but nonsignificant CIs for the association between Trust and glycemic control (OR, 0.78; 95% CI, 0.54-1.13). Sensitivity analyses with different cutoffs for SBP and LDL cholesterol control did not alter the results.
Using a novel, validated instrument to measure 3 aspects of culturally competent care in an ethnically diverse sample of low-income patients, we found that patient report of trust in their primary care physician was associated with better glycemic control. To our knowledge, this is the first US study providing empiric support for a relationship between patient trust in physician and a major diabetes outcome.23,24
Patient trust has been posited as central to effective culturally competent care.25,26 Among patients with diabetes, trust has been shown to increase self-efficacy and medication adherence,23 and is associated with reports of less difficulty in completing diabetes specific tasks.27 Effective communication between patient and physician has multiple components and has been conceptualized through at least 2 lenses: one focuses on an observable set of distinct communication behaviors (ie, physician asking open-ended questions, or using measures of shared decision-making); the other lens has been termed “relational” and has emphasized the subjective experience of communication by the patient and by the physician.28 The relational lens, with its emphasis on trust, has been paramount in conceptual models of cultural competence.13,29–32 Our findings support the importance of relational aspects of care in glycemic control and suggest that increasing patient physician trust might lead to improved glycemic control, particularly among racial and ethnic minorities with diabetes.
Our finding that only glycemic control and not lipid or blood pressure control is associated with the Trust domain is not entirely surprising. A recent study on patient physician language barriers among Latinos with diabetes showed a relationship between language concordance (which is associated with increased patient trust33) and glycemic control8; however, a second study in the same population revealed no relationship between language barriers and lipid or SBP control. Lipid and blood pressure control may be relatively easier to achieve than glycemic control, as lipid and blood pressure control may depend more on medication adherence alone than glycemic control, which additionally requires patient adherence to multiple, complex self-management behaviors surrounding diet and exercise. Glycemic control also differs from blood pressure or lipid management in the prevalent use of injectable medications. Insulin use has been associated with culturally specific concerns, and may require substantial trust in physician before being accepted by patients.34 Fewer patients in our sample had poor lipid or blood pressure control than poor glycemic making small differences more difficult to detect.
We found that the domain of Doctor Communication-Health Promotion was inversely associated with glycemic control. This finding is not necessarily surprising, as physicians may be more likely to emphasize diet and exercise recommendations when a patient has poor glycemic control. In contrast, we found that the domain of Doctor Communication-Positive Behavior was not associated with glycemic control. This is somewhat unexpected, as elements of this domain include “listens to patients” and “explains things clearly,” which would seem central to a strong doctor patient alliance. Further research disentangling trust from reports about general communication may help clarify the importance of each in patient care.
Our study has several limitations. The cross-sectional design does not allow us to infer causality between patient reports of trust in physician and better glycemic control. Second, while our patient population is representative of safety net patients in the San Francisco Bay Area and Chicago, our results may not generalize to privately insured or better educated patients with diabetes. Third, our results depend on patient report and, while patient subjectivity is probably the gold standard for reports of communication clarity and trust, we are unable to confirm reports of specific Doctor Communication-Health Promotion behavior. Fourth, our results may not generalize to other languages or ethnic groups.
Our study also has several strengths. We used domains from an instrument to measure patient report of cultural competence that has been validated in this low-income population in both English and Spanish.16 Second, our patient population is relatively large and representative of many public hospital and community clinics throughout the country.35 Third, we had objective measures of glycemic, lipid, and blood pressure control. Finally, few studies in this population with low educational attainment have been able to capture detailed information on patient reports of clinical care.
Given the high and disproportionate burden of diabetes on low-income, minority patients, improving glycemic control in this population is crucial to reducing the burden of diabetes-related morbidity, mortality, and health care costs in the United States. Our findings highlight the importance of patient physician trust in clinical care. Although more general physician cultural competency training interventions have been disappointing in their clinical results among patient with diabetes,6 it remains to be seen whether interventions that focus on relational aspects of health care are more successful.
APPENDIX 1: 5 DOMAINS OF THE CONSUMER ASSESSMENT OF HEALTH CARE PROVIDERS AND SYSTEMS CULTURAL COMPETENCY ITEM SET: CAHPS-CC
1. Stewart MA. Effective physician-patient communication and health outcomes: a review. CMAJ. 1995;152:1423–1433
2. Heisler M, Bouknight RR, Hayward RA, et al. The relative importance of physician communication, participatory decision making, and patient understanding in diabetes self-management. J Gen Intern Med. 2002;17:243–252
3. Heisler M, Cole I, Weir D, et al. Does physician communication influence older patients’ diabetes self-management and glycemic control? Results from the Health and Retirement Study (HRS). J Gerontol. 2007;62:1435–1442
4. Heisler M, Faul JD, Hayward RA, et al. Mechanisms for racial and ethnic disparities in glycemic control in middle-aged and older Americans in the health and retirement study. Arch Intern Med. 2007;167:1853–1860
5. Piette JD, Schillinger D, Potter MB, et al. Dimensions of patient-provider communication and diabetes self-care in an ethnically diverse population. J Gen Intern Med. 2003;18:624–633
6. Sequist TD, Fitzmaurice GM, Marshall R, et al. Cultural competency training and performance reports to improve diabetes care for black patients: a cluster randomized, controlled trial. Ann Intern Med. 2010;152:40–46
7. Schillinger D, Piette J, Grumbach K, et al. Closing the loop: physician communication with diabetic patients who have low health literacy. Arch Intern Med. 2003;163:83–90
8. Fernandez A, Schillinger D, Warton EM, et al. Language barriers, physician-patient language concordance, and glycemic control among insured Latinos with diabetes: the Diabetes Study of Northern California (DISTANCE). J Gen Intern Med. 2011;26:170–176
9. Greenfield S, Kaplan SH, Ware JE Jr., et al. Patients’ participation in medical care: effects on blood sugar control and quality of life in diabetes. J Gen Intern Med. 1988;3:448–457
10. Kaplan SH, Greenfield S, Ware JE Jr. Assessing the effects of physician-patient interactions on the outcomes of chronic disease. Med Care. 1989;27(suppl 3):S110–S127
11. Eliminate Disparities in Diabetes: Office of Minority Health, 2005
12. A Comprehensive Framework and Preferred Practices for Measuring and Reporting Cultural Competency: A Consensus Report. 2009 Washington, DC National Quality Forum (NQF)
13. Bigby J Cross-Cultural Medicine. 2003 Philadelphia, PA American College of Physicians
14. Fernandez A, Schillinger D, Grumbach K, et al. Physician language ability and cultural competence. An exploratory study of communication with Spanish-speaking patients. J Gen Intern Med. 2004;19:167–174
15. Weech-Maldonado R, Carle AC, Weidmer B, et al. The Consumer Assessment of Healthcare Providers and Systems (CAHPS) Cultural Competency (CC) item set. Med Care
. 2012;50(suppl 2):S22–S31
16. Stern RJ, Fernandez A, Jacobs EA, et al. Advances in measuring culturally competent care: a confirmatory factor analysis of the CAHPS–CC in a safety-net population. Med Care
. 2012;50(suppl 2):S49–S55
17. Zweifler J, Hughes S, Lopez RA. Controlling for race/ethnicity: a comparison of California commercial health plans CAHPS scores to NCBD benchmarks. Int J Equity Health. 2010;9:4–12
19. Aron D, Pogach L. Transparency standards for diabetes performance measures. JAMA. 2009;301:210–212
20. Saaddine JB, Cadwell B, Gregg EW, et al. Improvements in diabetes processes of care and intermediate outcomes: United States, 1988–2002. Ann Intern Med. 2006;144:465–474
21. Spitzer RL, Kroenke K, Williams JB. Validation and utility of a self-report version of PRIME-MD: the PHQ primary care study. Primary Care Evaluation of Mental Disorders. Patient Health Questionnaire. JAMA. 1999;282:1737–1744
22. Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med. 2001;16:606–613
23. Lee YY, Lin JL. The effects of trust in physician on self-efficacy, adherence and diabetes outcomes. Soc Sci Med (1982). 2009;68:1060–1068
24. Alazri MH, Neal RD. The association between satisfaction with services provided in primary care and outcomes in Type 2 diabetes mellitus. Diabet Med. 2003;20:486–490
25. Pearson SD, Raeke LH. Patients’ trust in physicians: many theories, few measures, and little data. J Gen Intern Med. 2000;15:509–513
26. Thom DH, Wong ST, Guzman D, et al. Physician trust in the patient: development and validation of a new measure. Ann Fam Med. 2011;9:148–154
27. Bonds DE, Camacho F, Bell RA, et al. The association of patient trust and self-care among patients with diabetes mellitus. BMC Fam Pract. 2004;5:26–32
28. Saba GW, Wong ST, Schillinger D, et al. Shared decision making and the experience of partnership in primary care. Ann Fam Med. 2006;4:54–62
29. Brach C, Fraser I. Can cultural competency reduce racial and ethnic health disparities? A review and conceptual model. Med Care Res Rev. 2000;57(suppl 1):181–217
30. Betancourt JR, Green AR, Carrillo JE, et al. Cultural competence and health care disparities: key perspectives and trends. Health Aff (Project Hope). 2005;24:499–505
31. Betancourt JR. Cultural competence—marginal or mainstream movement? New Engl J Med. 2004;351:953–955
32. Paez KA, Allen JK, Beach MC, et al. Physician cultural competence and patient ratings of the patient-physician relationship. J Gen Intern Med. 2009;24:495–498
33. Schenker Y, Karter A, Schillinger D, et al. The impact of limited English proficiency and physician language concordance on reports of clinical interactions among patients with Diabetes: the DISTANCE Study. Patient Educ Couns. 2010;81:222–228
34. Ratanawongsa N, Wright SM, Vargo EM, et al. Challenges in primary care relationships: seeing it from both sides. Patient Educ Couns. 2011;85:40–45
35. Chew LD, Schillinger D, Maynard C, et al. Glycemic and lipid control among patients with diabetes at six US public hospitals. J Health Care Poor Underserved. 2008;19:1060–1075