|Home | About | Journals | Submit | Contact Us | Français|
Interpersonal trust is an important component of the patient-doctor relationship. Little is known about patients’ trust in the multiple providers seen when confronting serious illness.
To characterize breast cancer patients’ trust in their regular providers, diagnosing physicians, and cancer treatment team and examine whether high trust in one’s regular provider confers high trust to cancer physicians.
704 white, black, and Hispanic breast cancer patients, age 30 to 79, with a first primary in situ or invasive breast cancer who reported having a regular provider.
We measure trust in: (1) regular provider, (2) diagnosing doctors, and (3) cancer treatment team. Other variables include demographic variables, preventive health care, comorbidities, time with regular provider, time since diagnosis, cancer stage, and treatment modality.
Sixty-five percent of patients reported high trust in their regular provider, 84% indicated high trust in their diagnosing doctors, and 83% reported high trust in their treatment team. Women who reported high trust in their regular provider were significantly more likely to be very trusting of diagnosing doctors (OR: 3.44, 95% CI: 2.27–5.21) and cancer treatment team (OR: 3.09, 95% CI: 2.02–4.72 ). Black women were significantly less likely to be very trusting of their regular doctor (OR: 0.58, 95% CI: 0.38–0.88) and cancer treatment team (OR: 0.45, 95% CI: 0.25–0.80). English-speaking Hispanic women were significantly less trusting of their diagnosing doctors (OR: 0.29, 95% CI: 0.11–0.80).
Our results suggest that patients are very trusting of their breast cancer providers. This is an important finding given that research with other populations has shown an association between trust and patient satisfaction and treatment adherence. Our findings also suggest that a trusting relationship with a regular provider facilitates trusting relationships with specialists. Additional work is needed to increase interpersonal trust among black women.
Interpersonal trust is an important component of the patient-doctor relationship.1,2 In a study of over 7,000 adults, trust in the primary care physician was strongly associated with patient satisfaction and adherence to treatment within the primary care setting.3 Other work has shown associations between trust and a greater desire for shared decision-making, medical adherence among diabetes patients, and acceptance and adherence to therapy among HIV-infected individuals.4–8 Low trust is associated with disenrollment from primary care, less satisfaction with care, lower intentions to follow physician advice, and lower self-reported symptom improvement.9,10
Interpersonal trust is also important for cancer screening and treatment. Cancer patients identify trust as an important aspect of their care and individuals, and high trust in the primary care physician is associated with adherence to cancer screening.11–14 Trust in cancer physicians is associated with enrollment in investigational cancer treatment protocols.15 In a study of breast and colon cancer patients, trust in the primary care physician was related to stage at diagnosis; patients reporting greater trust were more likely to be diagnosed at an earlier stage.16
Studies of interpersonal trust, including studies of cancer patients, have typically assessed trust of primary care physicians. Trust in primary care physicians is critical given the pivotal role of primary care providers in preventive health care and disease detection.17 Importantly, breast cancer care often begins with a regular provider who performs clinical breast examinations, discusses breast abnormalities, and makes referrals for mammography or additional screening. However, to our knowledge, no studies have considered cancer patients’ trust across the physicians seen in the course of cancer detection and treatment. Given the importance of having a trusting relationship with a primary provider, we were interested in understanding the extent to which a patient’s level of trust in her primary physician would confer trust to the unknown physicians involved in diagnosing and treating her breast cancer. Thus, we consider (1) levels of trust in cancer patients’ regular providers, diagnosing doctors, and treating doctors; and (2) whether having a high degree of trust in one’s regular provider confers high levels of trust to the physicians seen for cancer care.
Data are from a population of white, black, and Hispanic breast cancer patients (N=1,766). Eligible female patients were 30 to 79 years of age at diagnosis, resided in Chicago, had a first primary in situ or invasive breast cancer, and self-identified as non-Hispanic white, non-Hispanic black, or Hispanic. All eligible newly diagnosed cases were ascertained at 1 of 56 hospitals in the greater Chicago area. Certified tumor registrars employed by the Illinois State Cancer Registry (ISCR) visited hospitals monthly and reviewed pathology records, the hospital tumor registry, or both, depending on the protocol at the individual hospital, to identify new cases. Information on patient race and ethnicity was sought from patient medical records when not available in the hospital tumor registry.
ISCR mailed a letter describing the study to each eligible patient 1 to 3 months after diagnosis (in both Spanish and English if ethnicity was Hispanic or unknown). If a patient did not respond by mail or telephone within 10 days, a second contact was attempted by mail, telephone, or both. Once contact was established, interested patients were placed in contact with the UIC Survey Research Laboratory, screened for eligibility, and scheduled for an interview. If the patient was not interested, the case was flagged for re-contact 2 months later, allowing the patient more time to adjust to her diagnosis. Nine hundred eighty-nine women agreed to participate (response rate: 56%).
The survey interviewer obtained written informed consent before the interview. Patients received $100 for their participation. Interviews were administered in English or Spanish using computer-assisted personal interview (CAPI) procedures. CAPI allows interviewers to conduct face-to-face interviews using a portable computer instead of pen and paper. The interview included questions pertaining to the process of discovery, diagnosis, and treatment of the patient's breast cancer, health care-seeking behaviors, and social networks. The interview and consent process took approximately 90 min. All study procedures were approved by the University of Illinois at Chicago Institutional Review Board.
Because we were interested in interpersonal trust across the cancer experience, we assessed women’s trust in their regular providers, the doctors responsible for providing a cancer diagnosis, and the cancer treatment team. Trust in regular provider is measured with the following question: “In general, I trust my health care professional to give me the best possible health care. Would you say this is… (1) always true; (2) mostly true; (3) half the time true; (4) sometimes true; (5) never true.” Patients were asked a single item question about trust in their diagnosing doctors: "In general, how much did you trust your doctors to do everything they could to correctly diagnose your breast cancer?” Responses were on a 4-point scale (1=a great deal, 2=somewhat, 3=not too much, or 4=not at all). Patients were also asked about trust in their treatment team: "Let me ask you how you feel in general about the care you received from the doctors, nurses, and technicians during your treatment. In general, how much did you trust these people to provide you with the best possible health care?” Responses were on a 4-point scale (1=a great deal, 2=somewhat, 3=not too much, or 4=not at all). To minimize data sparseness issues and to allow for comparison across the three trust measures, we dichotomized responses. The highest level of trust is the index characteristic, and all other levels were combined for comparison.
Respondent characteristics obtained at interview included race/ethnicity (non-Hispanic white, non-Hispanic black, and Hispanic). The Hispanic group was further divided into two categories based on interview language. Additional demographic information included age (<50; 50 to 59; 60 or older), marital status (not married/not living with a partner, or married/living with a partner), income (<$25,000; $25,000–$62,500; or> $62,500), and education (less than a high-school degree, high-school graduate, some college). Twenty-five cases were missing income data and were assigned an income category based upon their education and the associated income distributions in the sample. Health insurance was categorized as no insurance, public insurance, or private insurance. Preventive health care was measured by a count of how many of the following exams a woman completed in the 2 years preceding her cancer diagnosis: routine physical, clinical breast exam, and mammogram. This variable ranged from 0 (completed none of the exams) to 3 (completed all three exams). We also included the number of comorbidities a woman reported (0, 1, 2, or more), and the number of years a woman had been with her regular provider was included (<1 year; 1–5 years; 6–10 years; and>10 years).
Breast cancer stage at diagnosis (stage 0, 1, 2, 3, or 4) was obtained from the patient's medical records and was based upon pathological analysis of the tumor. The number of days elapsed between the date of diagnosis and the interview was collapsed into quartiles (<85 days; 85–108 days; 109–148 days; and >149 days). We include two dichotomous variables representing whether the woman had undergone chemotherapy or a partial/complete mastectomy.
We selected only those respondents who reported having a regular provider. A patient was defined as having a regular provider if she answered yes to the following question: “Think back to the time before the problem was discovered that turned out to be cancer. Around that time, did you have a doctor or health care person that you thought of as your own doctor, someone you went to regularly for care?” Eighty-six percent (N=854) of patients reported having a regular provider. After removing 39 cases that lacked data on all three trust measures and 111 cases that lacked cancer stage data, our final sample consisted of 704 women.
Descriptive statistics and chi-squared tests were used to summarize patient characteristics and trust. Logistic regression for binary dependent variables was used to assess the association of personal and disease characteristics with trust and the association of trust in one’s regular provider with trust in cancer physicians. Five models were estimated. First, we model regular provider trust by patient and disease characteristics. Second, we model diagnosing doctor trust by regular provider trust. Third, we model diagnosing doctor trust by regular provider trust and patient and disease characteristics. Fourth, we model treatment team trust by regular provider trust and trust in diagnosing doctors. Finally, we model treatment team trust by regular provider trust, trust in diagnosing doctors, and patient and disease characteristics. All analyses were conducted using Stata/SE 9.2.
As shown in Table 1, the majority of the women were married, had more than a high school education, and had private insurance. Most of the women reported no comorbidities, had been with their regular provider for at least a year, and had a routine physical, clinical breast exam, and mammogram in the 2 years prior to their cancer diagnosis. The high percentage of routine care is expected given that these are women who have a regular health care provider. The majority of our sample was diagnosed with stage 0, 1, or 2 cancer. Approximately one-third of the women had undergone a mastectomy; a slightly larger percentage underwent chemotherapy.
Overall, 65% of breast cancer patients reported always trusting their regular provider. Rates of high trust in the regular provider varied significantly by race—71% of whites reporting high trust in their regular provider compared with 57% of blacks (p≤0.01). Among Hispanic women, 76% of English speakers and 64% of Spanish speakers reported high trust in their regular doctor. Women who had undergone a mastectomy were significantly less likely to have high trust in their regular doctor (p≤0.01). Eighty-four percent of respondents reported having a great deal of trust in their diagnosing doctors. High trust in diagnosing doctors was most common among whites (90%) and blacks (82%). English-speaking Hispanic women (76%) and Spanish-speaking Hispanic women were least likely to report high trust in diagnosing doctors (76%, p≤0.01). Greater formal education was associated with higher trust in diagnosing doctors (p≤0.001). Eighty-three percent of women reported a great deal of trust in their treatment team. Among whites, 90% reported a great deal of trust in their treatment team. Among Spanish-speaking Hispanic women, 86% reported a great deal of trust in their treatment team. Blacks and English-speaking Hispanics were significantly less likely to report a great deal of trust in their treatment teams (75% and 76%, respectively, p≤0.001). Higher income was associated with higher trust in treatment team (p<0.05).
Patterns of trust across the three providers are shown in Table 2. Fifty-five percent of patients were very trusting of all doctors. Nearly one-third of the sample did not report the highest levels of trust in their regular doctors, but did report the highest level of trust in their diagnosing and/or treating physicians. Approximately 10% of those with high trust in their regular doctor “lost” trust, with 4% not reporting high trust in their diagnosing physician, 4% not reporting high trust in their treatment team, and 2% not reporting high trust in both their treating and diagnosing physicians. Less than 5% of the women did not report the highest level of trust on any of the three measures. These women were similar to other women on demographic characteristics, disease stage, and chemotherapy. However, women who did not report high trust in any physician were significantly more likely to have undergone a mastectomy (p<0.05; results not shown here).
Table 3 presents results of the logistic regression analyses of trust in the regular physician. Patient race was a significant predictor of trust in the regular provider; black women had significantly lower odds of reporting high trust than white women, controlling for all other patient and disease characteristics (OR: 0.58, 95% CI: 0.38–0.88). Time since diagnosis was associated with trust, as was having a mastectomy. Women with a mastectomy were significantly less likely to have high trust in their regular physician (OR: 0.67, 95% CI: 0.47–0.95). In models not shown here, we used continuous measures of age, education, and income; results were similar to Table 3. Additionally, location of regular provider (private office or clinic, public facility, or private hospital) was not associated with provider trust (results not shown).
Table 4 presents results for trust in diagnosing doctor. As shown in Model 1, women with high trust in their regular provider had a three-fold increase in the odds of reporting high trust in their diagnosing doctors (OR: 3.44, 95% CI: 2.27–5.60). Controlling for patient and disease covariates (Model 2) did not reduce the association between trust in regular doctor and trust in diagnosing doctor (OR: 3.59, 95% CI: 2.30–5.60). English-speaking Hispanic women were less likely than white women to report high trust in their diagnosing doctors (OR: 0.29, 95% CI: 0.11–0.80). In Table 5, we report the results for trust in treatment team. As shown in Model 1, women who reported high trust in their regular provider had a threefold increase in the odds of reporting high trust in their treatment team (OR: 3.09, 95% CI: 2.02–4.72). Women who reported high trust in their diagnosing providers had 4.47 times the odds of reporting high trust in their treatment team than women who did not report high trust in their diagnosing doctors (95% CI: 2.81–7.11). Model 2 controls for patient and disease characteristics; however, the strength of the association between trust in regular and diagnosing doctors and trust in one’s treatment team is not diminished. Black women were significantly less likely to report high trust in their treatment team than white women (OR: 0.45, 95% CI: 0.25–0.80).
This study examined interpersonal trust in the physicians seen across breast cancer diagnosis and treatment. We found a high level of trust in regular physicians—65 percent of respondents reported a great deal of trust. Trust in cancer care providers was even higher; over 80 percent of respondents reported high trust in their diagnosing and treating physicians. When we considered a variety of patient and disease characteristics, trust in one’s regular provider was the strongest predictor of trust in one’s cancer providers. Patient race also affected trust. Most notably, among black women, only 57 percent reported a great deal of trust in their regular doctors. Black women exhibited greater trust in their cancer providers—82 percent reported high trust in their diagnosing doctors, and 75 percent reported high trust in their treating doctors. Interestingly, having a mastectomy was associated with lower trust in one’s regular doctor and with having low trust across all physicians.
Although primary care providers play an important role in referrals, treatment decision-making, support, and end of life care, few studies have examined the role of primary care providers in cancer care.18–20 Our findings speak to the importance of a trusting primary care relationship for subsequent trust in specialists. We hypothesize that trust is conferred through the referral process, as referral difficulties are associated with lower trust.21,22 Because a doctor’s prescription is required for a screening or diagnostic mammogram, regular providers likely played a key role in referrals. Our results are consistent with other studies that report lower trust of health care providers among blacks.21,23–26 However, this is the first study to investigate blacks’ levels of trust across multiple physicians. Given blacks’ history of abuse within the medical system, we might expect blacks to be the least trusting of physicians providing invasive cancer treatment. Thus, black women’s tendency to report higher trust in their cancer care providers than in their regular providers was surprising.
One caveat to our findings is the likely association between trust and participation in the study interview. Thus, the least trusting women may not be represented in our sample. In some cases a woman’s regular doctor may handle tissue samples and communicate a cancer diagnosis to her. However, women in large urban areas such as Chicago are typically referred to a specialist for biopsy and diagnosis. Nonetheless, our measures of trust in regular and diagnosing doctors can be thought of as measures of trust in two separate aspects of care—pre-cancer health care and cancer diagnosis—which may be provided by the same person or separate people. As with all survey research, our results may reflect unmeasured factors that shape trust. Recall bias may be present in this cross-sectional survey. The high levels of trust could reflect general tendencies to be trusting. However, prior research indicates that trust in physicians is based on real health care experiences, such as past disputes, amount of contact with physicians, evaluations of the competence of physicians, and choice in selecting a physician.27–30
Additional work is needed to examine trust among patients with other diseases. In one existing study, breast cancer patients were more trusting than other patients, perhaps due to their higher socioeconomic status relative to the other patients or to the uncertainty and risk provoked by breast cancer, which make trust especially salient and necessary for cancer patients.1,29 Second, because this sample was drawn from a variety of health care institutions, the trust levels we report are not a reflection of a particular health care environment; however, future work can examine how health care settings shape patient trust. Third, additional work is needed to examine how the process of cancer diagnosis and referral shapes trust. Fourth, additional research should examine how the content and quality of doctor-patient interactions and race-concordant patient-physician interactions shape trust.31,32 Finally, our findings suggest that efforts to improve trust among black women should focus on primary care. As the point where the breast cancer diagnosis and referral process begins, primary care physicians have a unique opportunity to instill trust in cancer patients. The role of primary care physicians in cancer care could increase as health care is restructured or as they assume a greater role in coordinating care, facilitating communication between providers, and maintaining patient awareness and empowerment.33 Trust of physicians among blacks may be augmented by improving physician interpersonal and technical competence and reducing black patients' expectations of racism and medical experimentation.34
This work was funded by the National Cancer Institute (Grant no. 5 P50 CA 106743) to the University of Illinois at Chicago Center for Population Health and Health Disparities, National Institutes of Health Grants R01 AG033172-01 and 1R21HD057473-01A1, and National Cancer Institute Training Program R25TCA57699-14. This work was presented at the annual meeting of the American Society of Preventive Oncology, March 2010.
Conflicts of interest None disclosed.