A minority of patients perceived having low care coordination (16.9%) or reported low satisfaction with care coordination (12.5%). There was a correlation coefficient of 0.47 between the two outcome variables. While 48.2% of those perceiving low coordination also reported low satisfaction, 51.8% of those perceiving low coordination did not report low satisfaction with these services. provides the univariate statistics for both measures of care coordination, including the means and standard deviations for the individual items used in the perceived care coordination scale.
Description of outcome measures (N=2,148)
3.1. Patient characteristics
Using sampling weights, the sample was 67.4% white, 14.8% African American, 7.9% Latina-highly acculturated and 8.8% Latina-less acculturated (1.1% were of other race/ethnicity). Just over 15% of respondents had low health literacy, 28.6% had moderate levels of health literacy and most (55.5%) had high health literacy.
Most patients had relatively high levels of educational attainment with two-thirds (65%) having some college or a college degree. The mean age was 57 (SD 11.3). About half the patients (50.6%) reported an annual family income of over $50,000, and most had some type of insurance. Respondents were diagnosed with varying stages of cancer, but most had invasive (Stage I–III) breast cancer (76.6%). While 44.1% of women reported none of the chronic conditions presented, 28.4% reported having one and 27.4% reported two or more.
The bivariate analyses revealed that minority patients more often reported perceiving poor care coordination than their white counterparts (22.1% of less acculturated Latinas, 20.2% of highly acculturated Latinas, 17.2% of African Americans and 14.4% of white patients, P=0.17)), but there were no significant differences in the reports of satisfaction with care coordination by race/ethnicity. Patients with low subjective health literacy more often reported low perceived care coordination and low satisfaction with care coordination ().
Care coordination perceptions by level of health literacy
Bivariate analyses also found some demographic factors (lower educational attainment, lower income and having no insurance) were significantly (P<0.05) associated with perceptions of low care coordination, while lower income and more comorbid conditions were associated (P<0.05) with low satisfaction with care coordination. Tumor stage was not associated with either measure of care coordination in the bivariate analysis. (results not shown).
3.2. Factors associated with patients’ perception of care coordination
shows the logistic regression results for factors associated with low perceived care coordination. The table displays the results for all independent variables included in the model. All other factors included in the model that were not significantly associated with perceived care coordination are indicated in the footnote to the table.
Logistic regression: Factors associated with perceptions of low care coordination among racially/ethnically diverse breast cancer patients (N=2,148)
As seen in Model 1, race/ethnicity was not significantly associated with perception of low care coordination. Age was inversely associated with perceptions of low care coordination; older women were less likely to perceive low care coordination than younger women (OR: 0.98; 95% CI: 0.97–0.99). Educational attainment was also inversely associated with perceptions of care coordination; women with higher levels of education were less likely to perceive low care coordination than those with lower educational attainment (OR=0.64; 05% CI: 0.48–0.85). No other patient factors were associated with perceptions of care coordination.
Model 2 shows that the addition of health literacy to the model did not substantially change the results of Model 1. The association between education and perception of low care coordination was slightly attenuated by the inclusion of subjective health literacy, but remained statistically significant at p<0.05. Health literacy was independently and significantly associated with perception of low care coordination. Patients with moderate or low health literacy were significantly more likely to report low care coordination than those with high health literacy. In particular, those with low health literacy were almost 4 times as likely as those with high health literacy to report low care coordination (OR=3.88; 95% CI: 2.78–5.41).
3.3. Factors associated with patients’ satisfaction with coordination of care
shows the same logistic regression models for low satisfaction with care coordination, with inclusion of all independent variables. Model 1 indicates that, as with perceptions of low care coordination, race/ethnicity was not significantly associated with low satisfaction with care coordination. Age was similarly inversely associated with satisfaction with care coordination; older women less often reported low satisfaction than younger women (OR=0.98; 95% CI: 0.97–0.99). In this model, having two or more comoribid conditions, compared to no comorbid conditions, was associated with low satisfaction with care coordination (OR=1.46; 95% CI: 1.02–2.11).
Logistic regression: Factors associated with low satisfaction with care coordination among racially/ethnically diverse breast cancer patients (N=2,148)
Model 2 shows that adding health literacy did not appreciably affect the associations observed in Model 1. As observed in the perceived low care coordination models (), patients with moderate or low subjective health literacy were significantly more likely than those with high health literacy to report low satisfaction with their care coordination. Again, those with low health literacy were substantially more likely than those with high health literacy to report low satisfaction (OR=3.19; 95% CI: 2.25–4.52)
Description of the patient sample (N=2,148)