Sociodemographic Characteristics of Midlife and Older Adults With Diabetes and Cardiovascular Conditions
The majority of diabetes and cardiovascular condition respondents were non-Hispanic Whites (51.3 percent diabetes, 62.2 percent cardiovascular conditions), followed by smaller proportions of non-Hispanic Blacks (19.7 percent diabetes, 17.0 percent cardiovascular conditions) and Hispanics (23.4 percent diabetes, 15.3 percent cardiovascular conditions) (Table 4). Respondents reporting "other"; or mixed race (10.7 percent diabetes, 5.4 percent cardiovascular conditions) were excluded from this analysis because many comparisons between men and women in "other"; race/ethnicity categories had too few observations to make reliable estimates, and the patterns did not appear to differ for younger vs. older women. Higher proportions of non-White minorities were Medicaid recipients, had public insurance only and did not attend college. More females than males reported these characteristics within each race/ethnicity category. Among non-White racial/ethnic minorities, more Medicare beneficiaries were also Medicaid recipients, compared to pre-Medicare midlife adults, although this trend was reversed in non-Hispanic White older adults. In general, more Medicare beneficiaries did not attend college, compared to pre-Medicare midlife adults. More Hispanics reported being uninsured than non-Hispanic Whites or non-Hispanic Blacks. These patterns were consistent among both respondents with diabetes and
Tables 5 and 6 summarize the relative rates of reported quality measures based on gender and age group comparisons for diabetes and cardiovascular conditions; Tables 7 and 8 show the relative rates comparing race/ethnicity groups for diabetes and for cardiovascular conditions. Actual measure rates are shown in Tables 9 and 10. A relative rate of 1 means the rates are the same for the two groups being compared; differences in bold indicate a statistically significant difference in reported quality of care measures (p<.05).
Comparisons in Quality of Care Based on Gender
For patients with diabetes, there were no statistically significant differences in performance between men and women for diabetes care, and only one for preventive care (i.e., non-Hispanic White men >65 are more likely to receive dental care compared to women—with a relative rate of 1.26) (Table 5). Non-Hispanic White men were less likely to report delays or unmet needs in care than non-Hispanic White women. The relative rate was 0.61, with 18.8 percent of non-Hispanic White men 45–64 reporting delay in getting necessary medical care compared to 31.0 percent of non-Hispanic White women in the same age group. This difference was smaller and was not significant for non-Hispanic White women of Medicare age (13.1 percent for men vs. 15.9 percent for women). There were few statistically significant differences in patient-centeredness items
Among respondents with cardiovascular conditions, gender comparisons tended to show advantages for women in the 45–64 age groups, with several comparisons achieving statistical significance (Table 6). Among Hispanics 45–64, 91.2 percent of men vs. 95.1 percent of women reported a blood pressure check during the year; among non-Hispanic Blacks 45–64, 64.4 percent of men vs. 71.0 percent of women received advice on exercise. This pattern did not hold for the >65 age group, where the relative rates tended to favor men or were not significant. Results for access to care and patient-centeredness were consistent with patterns in diabetes. Men reported fewer problems with access to care than women among respondents with cardiovascular conditions, and the differences were greater among the 45–64 age group than among the >65 age group. There were few significant gender differences in reports of patient-centered care.
Comparisons in Quality of Care Based on Race/Ethnicity
As shown in Table 7, non-Hispanic Whites were usually more likely than non-Hispanic Blacks or Hispanics to receive diabetes care, a flu vaccination and dental care. For example, non-Hispanic White men <65 were 33 percent more likely to receive an HbA1c test and an eye and foot examination, compared to non-Hispanic Black men in the same age group (44.0 percent of Whites and 33 percent of Blacks received all three services). The pattern was similar and usually significant for Hispanics. The relative rates for receiving a flu shot ranged from 1.18 for Hispanic women ≥65 to 1.48 for non-Hispanic Black women ≥65, compared to White women in the same age group. Among cardiovascular conditions, results were less consistent (Table 8). For men in both age groups, non-Hispanic Whites were more likely to receive advice on nutrition compared to non-Hispanic Blacks. For women ≥65, non-Hispanic Whites were less likely to receive advice on nutrition than non-Hispanic Blacks, and Hispanics ≥65 (both men and women) were also less likely to receive nutrition advice.
Compared to non-Hispanic Blacks and Hispanics, non-Hispanic Whites tended to report more problems in access to care, but for the most part, differences were not significant for either diabetes or cardiovascular condition patients. Similarly, there were few significant differences in patient-centeredness based on race/ethnicity. In general, non-Hispanic Whites tended to report better communication compared to non-Hispanic Blacks or Hispanics, but differences were small.
Comparisons in Quality of Care Based on Age
As shown in Table 5, respondents 45–64 with diabetes and cardiovascular conditions more often reported poor quality of care than their counterparts >65. Receipt of recommended diabetes care was lower among the 45–64 age group for all diabetes services among Hispanic men and women. For example, among Hispanic men, 89.8 percent of the 45–64 age group vs. 98.4 percent of the ≥65 age group received a cholesterol test, resulting in a relative rate of 0.91. Lower rates of some tests were also observed for non-Hispanic Whites and non-Hispanic Blacks. The 45–64 age group was less likely to receive flu shots but more likely to receive dental services. The lower rate of flu shots was expected because recommendations for this service sometimes begin at 50, particularly in times of vaccine shortage (Mardon et al 2006). For cardiovascular care, patients in the 45–64 age group were less likely to receive blood pressure checks but more likely to receive counseling on nutrition and exercise, compared to their older counterparts (Table 6).
Among people with diabetes and cardiovascular conditions, respondents 45–64 were significantly more likely to report delays in care; this was consistent across race/ethnicity and gender groups. For example, among respondents with cardiovascular conditions, the relative rates for the Unmet Needs composite ranged from 1.22 for non-Hispanic Black men to 1.91 for Hispanic men (Table 6). In the latter group, 13.8 percent of Hispanic men 45–64 vs. 7.2 percent of Hispanic men ≥65 answered "yes"; to the four items concerning unmet needs in care (Table 8). Respondents 45–64 also reported more problems on the
Figures 2–4 illustrate the complexity of results by focusing on several key measures of effectiveness, access to care and patient-centeredness, based on age, gender and race/ethnicity.
- Across all quality of care measures, there were few statistically significant gender differences. Differences tended to be small and did not favor women or men consistently. The only exception was in access to care, where women reported delays in care more frequently than men.
- Differences based on race/ethnicity were prominent for effectiveness measures. In general, non-Hispanic Whites had more positive outcomes, particularly for diabetic care, flu shots and dental care. Interestingly, non-Hispanic Whites were more likely to report problems in access to care. These differences are consistent across age groups. Patterns were not as consistent for cardiovascular conditions.
- There were consistent, significant, age-based differences in access to care.
- Although there were statistically significant differences in patient-centeredness measures, their effect was generally small and inconsistent across gender and race/ethnicity groups.