J Community Health (2010) 35:365–374 DOI 10.1007/s10900-010-9259-0
ORIGINAL PAPER
Diabetes Care and Outcomes: Disparities Across Rural America Nathan L. Hale • Kevin J. Bennett Janice C. Probst
•
Published online: 14 April 2010 Ó Springer Science+Business Media, LLC 2010
Abstract We examined differences in receipt of diabetes care and selected outcomes between rural and urban persons living with diabetes, using nationally representative data from the 2006 Behavioral Risk Factor Surveillance System (BRFSS). ‘‘Rural’’ was defined as living in a non-metropolitan county. Diabetes care variables were physician visit, HbA1c testing, foot examination, and dilated eye examination. Outcome variables were presence of foot sores and diabetic retinopathy. Analysis was limited to persons 18 and older self-reporting a diagnosis of diabetes (n = 29,501). A lower proportion of rural than urban persons with diabetes reported a dilated eye examination (69.1 vs. 72.4%; P = 0.005) or a foot examination in the past year (70.6 vs. 73.7%; P = 0.016). Conversely, a greater proportion of rural than urban persons reported diabetic retinopathy (25.8 vs. 22.0%; P = 0.007) and having a foot sore taking more than four weeks to heal (13.2 vs. 11.2%; P = 0.036). Rural residence was not associated with receipt of services after individual characteristics were taken into account in adjusted analysis, but remained associated with an increased risk for retinopathy (OR = 1.20, 95% CI = 1.02–1.42). Participation in Diabetes Self-Management Education (DSME) was positively associated with all measures of diabetes care included in the study. Availability of specialty services
N. L. Hale (&) J. C. Probst Department of Health Services Policy and Management, Arnold School of Public Health, South Carolina Rural Health Research Center, University of South Carolina, 220 Stoneridge Drive, Suite 204, Columbia, SC 29208, USA e-mail:
[email protected];
[email protected] K. J. Bennett Department of Family and Preventive Medicine, University of South Carolina School of Medicine, 3209 Colonial Drive, Columbia, SC 29203, USA
and travel considerations could explain some of these differences. Keywords Rural Diabetes care Diabetes outcomes Diabetes self management education
Introduction As the number of persons living with diabetes continues to increase [1], providing care necessary to reduce associated morbidity and mortality will become increasingly important. Diabetes care requires screenings, preventive services, self-management education and counseling be integrated with primary care services for treating routine conditions [2]. Recommendations outlining care all persons living with diabetes should receive have been developed by the American Diabetes Association (ADA) and include routine physician visits, Hemoglobin A1c (HbA1c) testing, foot examinations by a health professional and dilated eye examinations [3]. These services are important for prevention and early detection of complications associated with diabetes, including retinopathy and ulcerated foot sores. Studies using nationally representative survey data have tracked changes in diabetes care over time and disparities in care among certain sub-populations [4–8]. However, few have examined differences in diabetes care differed based on residence in rural areas. Meeting ADA recommendations for diabetes care can be challenging under optimal conditions, even more so for rural areas lacking the infrastructure to sustain processes needed to improve care and outcomes among persons living with diabetes [2]. Rural populations often lack adequate access to primary care and specialty care services [9], which are critical for providing
123
366
quality diabetes care. Studies conducted among rural or underserved populations noted difficulties in meeting the recommendations for diabetes care; however, results are localized and do not produce nationally representative estimates on differences in diabetes care among rural populations [10–14]. The ADA recommends that all individuals diagnosed with diabetes participate in Diabetes Self-Management Education (DSME). The effectiveness of DSME in improving diabetes care has been demonstrated [7, 15–17]; however, studies of participation in DSME among rural populations have noted significant challenges related to the availability and sustainability of DSME [18–20]. The extent to which these relationships impact diabetes care has not been fully explored. The purpose of the present study is to explore differences in diabetes care and selected outcomes associated with rural residence. Given the challenges rural populations face with access to adequate health care resources and utilizing DSME, it is believed that persons living with diabetes in rural areas will be less likely to receive recommended diabetes care than those residing in urban areas, resulting in worse outcomes.
Methods Design, Data Source and Sample We conducted a cross-sectional analysis of data from the 2006 Behavioral Risk Factor Surveillance System (BRFSS). The BRFSS is a state based survey system, coordinated and centrally compiled by the Centers for Disease Control and Prevention, which gathers data on preventive health practices, risk behaviors, and health care access. The BRFSS uses a multistage cluster design based on random-digit dialing to collect data on non-institutionalized civilian residents of the United States aged 18 years or older. Data from each state are pooled and weighted to produce nationally representative estimates. An optional module related to diabetes was completed by 47 states and included in the 2006 BRFSS dataset. Diabetes status was determined by response to the question of ‘‘Have you ever been told by a doctor you have diabetes?’’ Those responding ‘‘Yes’’ were considered to be persons living with diabetes and constitute the study sample of (n = 29,501). Once identified as having diabetes, individuals were asked additional questions on services, outcomes and disease management. Environmental data, such as rural versus urban residence, were drawn from Area Resource File (ARF). The ARF is sponsored by the Health Resources and Services Administration, USDHHS and compiles county level data pertaining to health facilities, health professionals, health
123
J Community Health (2010) 35:365–374
outcomes and basic demographics. The ARF data were linked to the BRFSS based on respondents’ reported county of residence. Independent Variable: Rural Residence Definitions of rural were based on the 2003 Urban Influence Codes (UICs) from the United States Department of Agriculture Economic Research Service [21]. UICs divide the 3,141 counties, county equivalents, and independent cities into 12 groups based on population and commuting data from the 2000 Census [22]. For the purposes of this study, UICs of 1 and 2 were classified as ‘‘Urban,’’ while UICs 3–12 were considered ‘‘Rural.’’ Rural counties vary in size, and may contain modest urbanized areas of less than 50,000 population. Dependent Variables Selected services and outcomes associated with diabetes care were used to examine differences in care based on residence. Having one or more physician visits in the past 12 months was the first dependent variable of interest and was coded dichotomously (yes/no). Values for HbA1c testing were derived from the question ‘‘How many times in the past 12 months has a doctor, nurse, or other health professional checked for hemoglobin ‘A one C’.’’ Individuals reporting two or more HbA1c tests in the past 12 months were considered to meet ADA recommendations [3]. The foot exam variable was derived from the question ‘‘About how many times in the past 12 months has a health professional checked your feet for any sores or irritations?’’ Persons reporting their feet were checked at least once in the past 12 months were considered to meet recommendations [3]. Values for eye exams were derived from the question ‘‘When was the last time you had eye exam in which the pupils were dilated? This would have made you temporarily sensitive to bright light?’’ Individuals reporting at least one dilated eye exam in the past 12 months were considered to meet recommendations [3]. Two self-reported diabetic sequellae were also studied. Persons with diabetes were asked ‘‘Has a doctor ever told you that diabetes has affected your eyes or that you had retinopathy?’’ Persons responding ‘‘yes’’ were considered to have diabetic retinopathy. The foot sore variable was derived from responses to the question ‘‘Have you ever had any sores or irritation on your feet that took more than four weeks to heal?’’ and was also coded dichotomously (yes/no). Covariates Anderson’s Behavioral Model for Health Services Use served as the conceptual framework for defining covariates
J Community Health (2010) 35:365–374
that might affect the receipt of diabetes care and diabetes outcomes [23]. The Anderson model conceptualizes health utilization as a function of the individual’s predisposing characteristics such as age, enabling factors such as insurance coverage, and need for services, together with environmental factors such as the availability of services. Predisposing characteristics held constant in the analysis include sex, race/ethnicity, and age. Race/ethnicity was a four level variable, non-Hispanic White, non-Hispanic Black, non-Hispanic Other and Hispanic. Age was grouped into ages of 18–34, 35–54, 55–64 and 65 and older. Enabling factors available in the 2006 BRFSS included education, marital status, income, participation in DSME, insurance status, and not seeing a provider due to cost. Education was coded as high school diploma or less, some college or technical school and college graduate or above. Marital status was coded as married, divorced or separated, and other. Values for income were \ $25,000, $25,000– $50,000, and [$50,000.Values for participation in DSME were derived from the question ‘‘Have you ever taken a course or class in how to manage your diabetes yourself?’’ Health insurance status was derived from the question ‘‘Do you have any kind of health care coverage, including health insurance, prepaid plans such as HMOs, or government plans such as Medicare?’’ and was coded dichotomously as (yes/no). Not seeing a provider due to cost was taken from the question ‘‘Was there a time in the past 12 months when you needed to see a doctor but could not because of cost?’’ and was also coded dichotomously. Factors of individual need recorded by the BRFSS include insulin dependence, overweight or obesity status, smoking status, and self-reported health status. The overweight/obese variable was derived from the respondent’s Body Mass Index (BMI) and defined as greater than 25, or above the 85th percentile. Values for self-reported health were coded dichotomously; Excellent/Very Good/Good Health versus Fair/Poor Health. Because rural counties are not homogenous, environmental characteristics obtained from the ARF were also included as covariates. Health Professional Shortage Area (HPSA) designation serves as an important measure of primary care physician availability in a given county. Counties can be designated as a whole or partial HPSAs based on primary care physician to population ratio and primary care service need. Counties or partial counties with a physician to population ratio of at least 3,500 to 1 are considered HPSAs. In addition, those with a physician to population ratios greater than 3,000 to 1, but with high needs for primary care services or insufficient capacity of existing primary care providers are considered a HPSA [24]. Values for HPSA designation were coded as a 3 level categorical variable; not a HPSA, partial county HPSA and whole county HPSA.
367
Statistical Methods Bivariate analysis was conducted to compare outcomes among categorical variables using Chi-squared tests for independence. In addition, unadjusted odds ratios were obtained for rural and urban differences in recommended services and selected outcomes. Multivariate logistic regression was performed with each of the recommended services and selected outcomes serving dependent variable. Rural residence was the primary independent variable of interest while the described individual and environmental characteristics were included as covariates in each model. To account for the multistage, complex sampling design of the BRFSS, all analysis was conducted using SAS-callable SUDAAN (RTI International, Research Triangle Park, NC).
Results Characteristics of Population The demographic characteristics of adults with diabetes reached by the 2006 BRFSS are shown in Table 1, stratified by residence. A higher proportion of rural than urban persons reported diabetes among all racial/ethnic classifications (9.0 vs. 7.7%; P B 0.001). Rural non-Hispanic Blacks had the highest proportion of diabetes among all racial/ethnic and geographic categories (13.4%). Characteristics of persons with diabetes are shown in Table 2. The population of persons with diabetes in rural areas has a slightly higher proportion of women (50.6 vs. 47.6%; P = 0.01). Rural residents with diabetes were disadvantaged compared to urban residents in several important enabling areas, including lower levels of educational attainment and a greater likelihood of being in the lowest income category (Table 2). Thus, 58% of rural residents, versus
Table 1 Proportion of adults reporting diabetes, by residence and race/ethnicity, 2006 BRFSS Residence Urban (n = 16,406)
Rural (n = 6,695)
\0.001
Race All
7.7
9.0
White
7.0
8.7
Black
11.8
13.4
8.0 7.7
10.9 6.8
Other Hispanic
P value
All results are reported in weighted percentages, P value indicates rural/urban differences
123
368
J Community Health (2010) 35:365–374
Table 2 Characteristics of adults reporting diabetes, by residence, 2006 BRFSS
Residence Urban (n = 16,406)
Rural (n = 6,695)
Female
47.6
50.6
Male
52.4
49.4
White
85.06
14.94
Black
89.77
10.23
Other
88.73
11.27
Hispanic
96.11
3.89
P value
Predisposing Gender
0.014
\0.001
Race
Age
0.974
18–34 35–54
4.7 29.6
5.0 29.3
55–64
26.4
26.4
65?
39.2
39.3
49.3
58.4
Enabling \0.001
Education High school grade or less Some college
25.4
23.7
College graduate
25.3
17.9
Marital status
0.010
Married
58.9
62.1
Divorced
15.00
14.6
Other
26.1
23.3
\$25,000
38.6
47.7
$25,000–$50,000 [$50,000
29.6 31.8
29.3 23.1
55.8
52.0
\0.001
Income
Participation in DSME Health insurance (percent uninsured)
9.1
11.3
0.009
12.8
17.0
\0.001
Insulin dependent (percent yes)
25.7
26.9
0.298
Overweight or obese (percent yes)
83.4
85.5
0.017
Current smoker (percent yes)
15.5
17.2
0.070
Good to excellent
53.5
49.5
Fair to poor
46.5
50.5
No HPSA
14.3
25.5
Partial county HPSA
84.5
65.1
Whole county HPSA
1.2
9.4
Not seeing provider due to cost Individual need
\0.001
Self reported health status
Environment \0.001
HPSA status All results are reported in weighted percentages, P value indicates rural/urban differences
49% of urban, had only a high school education. Also, rural persons with diabetes were more likely to lack health insurance and to report not seeing a physician due to cost, and less likely to report participation in DSME than urban persons. From a family support perspective,
123
rural residents were more likely to be married than their urban peers. While rural and urban residents did not differ regarding insulin dependence, a higher proportion of rural persons with diabetes were overweight/obese (85.5%) compared to
J Community Health (2010) 35:365–374
369
their urban counterparts (83.4%; P = 0.02). Fewer rural persons with diabetes reported good to excellent health compared to urban persons (49.5 vs. 53.5%; P B 0.001). Among environmental factors influencing diabetes care, persons living with diabetes residing in rural areas were more likely to be in a whole county HPSA than urban persons (9.4 vs. 1.2%; P B 0.001). Diabetes Services Receipt of selected health services and outcomes are shown in Table 3, stratified by residence. No significant differences were observed between rural and urban persons with diabetes regarding a physician visit in the past 12 months (88.8 vs. 89.5%; P = 0.513) or reporting at least two HbA1c tests in the past year (66.6 vs. 67.3%; P = 0.588). The proportion of rural residents reporting an annual dilated eye examination (69.1%), however, was significantly lower than among urban persons with diabetes (72.4%; P = 0.006). Rural persons with diabetes were also less likely to report having an annual foot examination by a health professional in the past year than their urban counterparts (70.6 vs. 73.7%; P = 0.016). Adjusted Analysis of Receipt of Diabetes Services Physician Visit As shown in Table 4, after adjusting for enabling factors, predisposing characteristics and individual measures of need, there were no significant differences in the odds of reporting a physician visit based on rural residence (OR = 0.90, 95% CI = 0.75–1.09). Individuals residing in a partial county HPSA, however, were less likely to have a physician visit than those not residing in a HPSA (OR = 0.76, 95% CI = 0.62–0.93). Those participating in
DSME were more likely to have a physician visit in the past year (OR = 1.65, 95% CI = 1.35–2.03). Individuals disadvantaged by lack of health insurance or previously deferring care due to cost were less likely to report a physician visit. Income, however, showed a curvilinear distribution, with persons in the middle-income category (between $25,000 and $50,000) but not those in the lowest category being less likely to report a visit (OR = 0.77, 95% CI = 0.60–0.99). With enabling factors held equal in analysis, non-Hispanic Blacks (OR = 1.46, 95% CI = 1.04–2.06) and nonHispanic Others (OR = 1.82, 95% CI = 1.12–2.94) were more likely than whites to report a physician visit in the past year. Need was markedly associated with reported physician visit, with persons reporting insulin dependence (OR = 3.66, 95% CI = 2.76–4.87) and those who were overweight/ obese (OR = 1.39, 95% CI = 1.02–1.89) being more likely to report a physician visit than their counterparts.
HbA1c Testing Rural residence was not significantly associated with the likelihood of reporting HbA1c testing in adjusted analysis (OR = 1.06, 95% CI = 0.93–1.20), nor was HPSA status of the county. Persons with diabetes participating in DSME were more likely to have at least two HbA1c tests in the past year (OR = 1.72, 95% CI = 1.48–1.99). Paralleling results for physician visits, the odds of receiving at least two HbA1c tests in the past year was reduced among persons lacking health insurance (OR = 0.60, 95% CI = 0.45–0.79), not seeing a provider in the previous year due to cost (OR = 0.55, 95% CI = 0.44–0.69), and having an income between $25,000 and $50,000 (OR = 0.77, 95% CI = 0.60–0.99). Among individual measures of need, being insulin dependent (OR = 1.78, 95% CI = 1.49–2.12) was
Table 3 Diabetes care and selected outcomes among adults living with diabetes by residence, 2006 BRFSS Percent receiving service or experiencing outcome, by location
Odds that a rural resident will receive services
P value
% Urban (referent)
% Rural
OR (LBL–UBL)
Provider visit in past 12 months
89.5
88.9
0.95 (0.80–1.14)
0.513
At least 2 Hba1c test in past year
67.3
66.6
0.97 (0.87–1.09)
0.588
At least 1 foot exam in past year
73.7
70.9
0.89 (0.79–1.00)
0.016
Dilated eye exam in past year
72.4
69.1
0.86 (0.77–0.97)
0.006
Reported retinopathy
22.0
25.8
1.21 (1.04–1.40)
0.007
Foot sore ([4 weeks to heal)
11.2
13.2
1.21 (1.02–1.43)
0.036
Service receipt
Outcomes
All results are reported in weighted percentages, P value indicates rural/urban difference
123
123 0.85 (0.74–0.99) 0.61 (0.43–0.86) 1.16 (0.95–1.42) 0.96 (0.78–1.17)
0.82 (0.67–0.99) 0.88 (0.73–1.07) 0.86 0.75 1.72 0.60 0.55
0.85 (0.69–1.05) 0.75 (0.44–1.28) 1.23 (0.95–1.60) 0.90 (0.66–1.21)
1.25 (0.97–1.60) 1.03 (0.79–1.33) 0.88 0.77 1.65 0.46 0.65 1.78 (1.49–2.12) 1.12 (0.92–1.36) 0.91 (0.75–1.11) 1.09 (0.94–1.28)
1.05 (0.90–1.24) 0.81 (0.60–1.11)
3.66 (2.76–4.87) 1.39 (1.02–1.89) 1.02 (0.79–1.31) 1.12 (0.90–1.38)
0.76 (0.62–0.93) 0.87 (0.56–1.35)
(0.68–1.08) (0.61–0.92) (1.48–1.99) (0.45–0.79) (0.44–0.69)
0.86 (0.69–1.08) 0.90 (0.66–1.23) 0.85 (0.64–1.14)
1.46 (1.04–2.06) 1.82 (1.12–2.94) 1.08 (0.71–1.65)
(0.65–1.20) (0.60–0.99) (1.35–2.03) (0.33–0.63) (0.50–0.85)
1.06 (0.93–1.20)
0.90 (0.75–1.09)
Included in the analyses but not shown because not significant for any outcome: marital status
Predisposing Residence (ref: urban) Rural Race/ethnicity (ref: white) Black Other Hispanic Gender (ref: female) Male Age (ref: 35–54) 18–34 55–64 65? Enabling Education (ref: college grad) \High school High school some college Income (ref: [ $50,000) \$25,000 $25,000–$50,000 Participation in DSME (ref: none) No health insurance (ref: insured) Not seeing provider due to cost in past year (ref: saw provider) Individual need Insulin dependent (ref: not insulin dependent) Overweight or obese (ref: normal weight) Current smoker (ref: non–smoker) Self reported health status (ref: good to excellent) Fair to poor Environment HPSA (ref: no HPSA) Partial county Whole county
Dilated eye exam OR (LBL–UBL)
(0.86–1.37) (0.74–1.11) (1.83–2.42) (0.56–0.94) (0.49–0.75)
1.00 (0.86–1.17) 0.99 (0.71–1.38)
1.16 (1.00–1.35)
2.19 (1.82–2.63) 1.14 (0.94–1.37) 1.01 (0.83–1.22)
1.08 0.91 2.10 0.72 0.61
0.90 (0.74–1.08) 0.94 (0.77–1.15)
0.84 (0.58–1.22) 1.24 (1.03–1.50) 1.41 (1.15–1.72)
1.16 (1.00–1.34)
1.73 (1.38–2.16) 1.04 (0.77–1.39) 1.38 (1.02–1.87)
0.95 (0.83–1.09)
(0.61–1.02) (0.68–1.04) (1.66–2.25) (0.53–0.93) (0.46–0.71)
1.01 (0.86–1.19) 0.81 (0.57–1.13)
1.00 (0.85–1.18)
1.60 (1.34–1.90) 1.15 (0.95–1.40) 0.77 (0.63–0.94)
0.79 0.84 1.93 0.70 0.57
0.80 (0.65–0.98) 0.81 (0.66–1.01)
0.77 (0.54–1.11) 1.39 (1.15–1.70) 2.32 (1.89–2.84)
0.89 (0.77–1.04)
1.16 (0.94–1.44) 1.27 (0.95–1.70) 1.00 (0.73–1.37)
0.96 (0.84–1.09)
(1.11–2.07) (1.00–1.72) (0.81–1.22) (0.59–1.44) (0.94–1.70)
0.85 (0.68–1.05) 1.14 (0.77–1.70)
1.88 (1.55–2.28)
3.40 (2.82–4.10) 0.91 (0.71–1.16) 1.00 (0.72–1.38)
1.52 1.31 1.00 0.92 1.26
0.91 (0.70–1.19) 0.84 (0.65–1.09)
0.55 (0.30–1.00) 1.39 (1.08–1.80) 1.51 (1.14–1.99)
1.32 (1.10–1.60)
1.80 (1.36–2.38) 1.48 (0.99–2.23) 1.89 (1.32–2.71)
1.20 (1.02–1.42)
Diabetic retinopathy OR (LBL–UBL)
Foot exam OR (LBL–UBL)
Provider visit OR (LBL–UBL)
At least 2 Hba1c Test OR (LBL–UBL)
Outcomes
Service receipt
Table 4 Odds that a person with diabetes will report receipt of selected services and selected outcomes, 2006 BRFSS
(1.43–3.01) (0.93–1.66) (0.88–1.40) (0.64–1.99) (0.96–1.95)
1.12 (0.91–1.39) 0.96 (0.66–1.39)
2.21 (1.71–2.86)
2.00 (1.62–2.48) 1.20 (0.91–1.59) 1.06 (0.82–1.38)
2.07 1.24 1.11 1.13 1.36
0.66 (0.51–0.84) 1.02 (0.75–1.38)
0.66 (0.40–1.08) 0.73 (0.55–0.97) 0.54 (0.40–0.73)
1.11 (0.88–1.39)
0.60 (0.45–0.82) 1.04 (0.72–1.51) 0.83 (0.55–1.25)
1.06 (0.87–1.28)
Foot sore OR (LBL–UBL)
370 J Community Health (2010) 35:365–374
J Community Health (2010) 35:365–374
associated with having at least two HbA1c tests in the past year. Annual Dilated Eye Exam In adjusted analysis, rural and urban residents did not differ in the probability of reporting an annual dilated eye exam in the past year (OR = 0.96, 95% CI = 0.84–1.09). Residing in a partial (OR = 1.01, 95% CI = 0.86–1.19) or whole (OR = 0.81, 95% CI = 0.57–1.13) county HPSA was also not significantly associated with having an annual dilated eye exam (OR = 0.94, 95% CI = 0.80–1.11). Individuals who participated in DSME were more likely to report an annual dilated eye exam than their counterparts (OR = 1.93, 95% CI = 1.66–2.25). Enabling factors reducing the likelihood of having an annual dilated eye exam included less than a high school education (OR = 0.80, 95% CI = 0.65–0.98), lack of health insurance (OR = 0.70, 95% CI = 0.53–0.93) and not seeing a doctor in the past 12 months due to cost (OR = 0.57, 95% CI = 0.46–0.71). Among predisposing characteristics, adults between the ages of 55–64 (OR = 1.39, 95% CI = 1.15–1.70) and those 65 or over (OR = 2.32, 95% CI = 1.89–2.84) were more likely to report having an annual dilated eye exam than those in the 35 – 54 age group. Need was also associated with having an annual dilated eye exam, those reporting insulin dependence (OR = 1.60, 95% CI = 1.34–1.90) were more likely to have an annual dilated eye exam. Being a smoker reduced the likelihood of having an annual dilated eye exam. (OR = 0.77, 95% CI = 0.63–0.94). Annual Foot Exam by Health Professional Adjusted analysis yielded no significant differences between rural and urban persons living with diabetes in the odds of reporting an annual foot exam (OR = 0.95, 95% CI = 0.83–1.09). Likewise, residing in a whole (OR = 1.00, 95% CI = 0.86–1.17) or partial (OR = 0.99, 95% CI = 0.71–1.38) county HPSA was not associated with having an annual foot exam. However, participation in DSME was associated with an increased likelihood of receiving an annual foot exam (OR = 2.10, 95% CI = 1.83–2.42). Other enabling factors reducing the likelihood of having a foot exam include having no health insurance (OR = 0.72, 95% CI = 0.56–0.94) and not seeing a provider due to cost (OR = 0.61, 95% CI = 0.49– 0.75). Among predisposing characteristics associated with diabetes care, non-Hispanic Blacks (OR = 1.73, 95% CI = 1.38–2.16) and those of Hispanic ethnicity (OR = 1.38, 95% CI = 1.02–1.87) were more likely to report a foot exam than non-Hispanic Whites. In addition, individuals
371
between the age of 55 to 64 (OR = 1.24, 95% CI = 1.03– 1.50) and those 65 and over (CI = 1.41, 95% CI = 1.15– 1.72) were more likely to report having a foot exam in the past year. Need was also associated with having a foot exam with those insulin dependent (OR = 2.19, 95% CI = 1.82–2.63) being more likely to have a foot exam than their counterparts. Diabetes Outcomes, Prevalence and Adjusted Analysis Diabetic Retinopathy As shown in Table 3, rural persons living with diabetes were more likely to indicate having retinopathy (25.8%) than were 22.0% of urban residents (OR 1.21; P = 0.007). This relationship was retained in adjusted analysis, as rural residents with diabetes were more likely to indicate having retinopathy than were their urban peers (OR = 1.20, 95% CI = 1.02–1.42). Neither residing in a whole or partial county HPSA, nor participating in DSME was associated with having diabetic retinopathy. While uninsured persons and those deferring care were at lower odds for receipt of services in the past year, these financial access characteristics did not affect the odds for retinopathy, although lowincome persons did have increased risk (Income under $25,000 versus over $50,000, OR = 1.252, 95% CI = 1.11–2.07). Minorities, specifically non-Hispanic Blacks (OR = 1.80, 95% CI = 1.36–2.38) and Hispanic adults (OR = 1.89, 95% CI = 1.32–2.71) had higher odds for retinopathy compared to non-Hispanic Whites Table 4). Males and older adults were also more likely to report retinopathy. Need factors associated with an increased likelihood of having retinopathy include being insulin dependent (OR = 3.40, 95% CI-2.89–4.10) and having self reported health of fair to poor (OR = 1.88, 95% CI = 1.55–2.28). Foot Sore with More than Four Weeks to Heal The proportion of rural persons indicating having a foot sore taking more than four weeks to heal was significantly greater than of urban persons (13.2 vs. 11.2%, OR 1.21, P = 0.036). In adjusted analysis, however, the relationship between residence and foot sores was attenuated (OR = 1.06, 95% CI = 0.87–1.28). Neither residing in a partial or whole county HPSA, nor participating in DSME was associated with having a foot sore taking more than four weeks to heal. Among enabling factors, reporting an income less than $25,000 was associated with an increased likelihood of having a foot sore taking more than four weeks to heal (OR = 2.07, 95% CI = 1.43–3.01). Among predisposing characteristics, non-Hispanic Blacks (OR = 0.60, 95% CI = 0.45–0.82), those with less
123
372
than a high school education (OR = 0.66, 95% CI = 0.51– 0.84), between the ages of 55–64 (OR = 0.73, 95% CI = 0.55–0.97) and those 65 or older (OR = 0.54, 95% CI = 0.40–0.73) were less likely to have a foot sore taking more than four weeks to heal. Need factors associated with an increased probability of having a foot sores include insulin dependence (OR = 2.00, 95% CI = 1.62–2.48) and a self reported health status of fair to poor (OR = 2.21, 95% CI = 1.71–2.86).
Discussion Our analysis confirmed that rural residents with diabetes differ from their urban counterparts. The prevalence of self-reported diabetes in adults is 17% higher in rural counties than in metropolitan areas (9.0 vs. 7.7%), across all race/ethnicity groups except Hispanics. The presence of rural disparities in diabetes prevalence echoes prior work using the 1988–1994 National Health and Nutrition Examination Survey [25]. Because rural residents suffer a higher burden from diabetes, they constitute an important public health target group. Rural adults with diabetes differed from their urban counterparts in ways that affect planning for rural interventions. Rural adults were less well educated, more likely to report low incomes, more likely to lack health insurance, and correspondingly, more likely to report deferring care due to cost than urban adults. In addition, other research has shown that rural residents travel further for care [26]. Despite these disadvantages, rural persons with diabetes were no less likely to report receiving two Hga1c tests in the past year than urban residents, although only about two thirds of persons in either geographic area met this guideline. However, a smaller proportion of rural than urban adults with diabetes reported receipt of a foot exam or a dilated eye examination. Thus, it is not surprising that slightly more rural than urban respondents noted foot sores of more than one month’s duration and a diagnosis of retinopathy. The proportion of rural persons reporting a dilated eye exam in the present study (68.1%), is slightly higher than previous population estimates produced using 2002 and 2003 BRFSS data [6, 8], while similar to that in research among Medicare beneficiaries for the 1999–2001 period [27]. Nonetheless, performance on this measure remains below the Health People 2010 goal of 75% [28]. On all service receipt variables, lack of health insurance and reported deferring of care due to cost were strongly associated with reduced odds for service receipt in adjusted analysis, while rural residence alone ceased to have an association with services. This suggests that programs focusing on reducing cost barriers may improve service quality on this measure.
123
J Community Health (2010) 35:365–374
Retinopathy is a common result of diabetes, but receipt of recommended care can lower the risk of vision loss [29] and slow the resulting decline in health related quality of life [30]. The reduced rate at which rural residents receive dilated eye examinations compared to urban adults appears attributable to lower education and income levels in the rural population, both of which reduced the likelihood of receiving an examination. However, the higher prevalence of diabetic retinopathy among rural adults was virtually unchanged in adjusted analysis (OR 1.21; AOR 1.20). With the exception of income, access variables (insurance, deferred care) were not significantly associated with diabetic retinopathy, while risk factors were markedly linked, including insulin dependence, perceived fair/poor health, and age of 55 or higher. It is possible that rural practitioners, who typically see more patients and provide fewer preventive screening services than urban physicians [31] are not sufficiently aggressive in educating and monitoring patients with diabetes. A recent study found that counties with rural health clinics, which are required to incorporate a midlevel practitioner in addition to one or more physicians, have better population based rates for eye screening than other rural counties [32]. Increasing use of non-physician workforce, therefore, may be one approach for improving eye outcomes among rural patients. The proportion of rural persons receiving a foot exam by a health professional in the past year was 70.9%, slightly less than among urban persons with diabetes (73.7%) and short of the Healthy People 2010 goal of 75% [28]. Since our study found no differences in the proportion of persons reporting a physician visit within the past year, and a foot examination does not require special equipment or access to specialty care, the shortfall is difficult to explain. With no differences in reported physician visits, it should be expected foot care and outcomes would not differ by residence. However, this is not the case. Efforts to implement quality improvement initiatives to improve diabetes care among rural and underserved populations are critical for increasing the number of persons receiving preventive services and improving outcomes. Certain quality initiatives among rural and underserved populations have show improvements in diabetes care and warrant consideration. A group of 19 Midwestern community health centers implemented Plan-Do-Study-Act cycles to improve diabetes care provided to patients. Significant improvements in HbA1c testing, referrals for eye examinations and foot examinations were noted after one year [33]. Other examples of collaborative efforts between multiple partners to improve diabetes care have also been noted. The Montana diabetes prevention and control programs worked with the University of North Dakota to support 37 rural primary care practices to improve diabetes care. This collaborative effort also demonstrated
J Community Health (2010) 35:365–374
improvements in HbA1c testing, eye and foot examinations among patients [2]. Participation in DSME was positively associated with quality of care measures, including a reported physician visit in the past year, two or more HbA1c tests in the past year, an annual dilated eye exam and an annual foot exam by a health professional. This finding is consistent with previous studies examining participation in DSME among the general population [7, 17]. However, rural residents were less likely to have participated in DSME, with 52.0% of rural persons living with diabetes participated in DSME compared to 55.9% of urban persons (P = 0.003), a finding that parallels previous research [13, 18, 20]. Given the strong positive relationship between participation in DSME and diabetes care, efforts should continue to ensure rural populations have adequate access and utilize DSME services. Various strategies to increase access to DSME among rural populations have shown promise. One state established a mentoring program to increase the number of Certified Diabetes Educators (CDE) in rural and frontier areas and increase the number certified education programs available in rural areas [19]. Others have implemented elements of a chronic care model by placing a CDE in a physician practice to support providers in diabetes management and provide patient education [34]. More recently, the expansion of telemedicine has created the opportunity to deliver DSME in rural areas. A Pilot project using telemedicine to deliver DSME demonstrated improved knowledge, self-efficacy and more frequent self care among participants [35]. The present study has several limitations. First, a cross-sectional study can only present associations among variables, not causal relationships. Next, all information is self-reported. Third, a broad definition of rural was used, potentially masking notable differences in diabetes care that may be present in very small or remote rural counties. Fourth, the study only considers the receipt of specific health services and does not account for the content or quality of care being provided, nor how well diabetes is being controlled through follow up activities. Next, participation in DSME demonstrated a strong positive relationship with diabetes care. However, individuals voluntarily participating in DSME might also be more inclined to take an active role in disease management, introducing selection bias. Finally, the BRFSS is a telephone survey of non-institutionalized adults; therefore, individuals such as those in nursing homes, or without a telephone could be excluded [17]. Rural residents as a whole are more likely to report having diagnosis, less likely to receive recommended services, and more likely to suffer from diabetic retinopathy. The characteristics of rural populations such as lower reported levels of income, educational attainment and health insurance, rather than location alone, place them at increased risk. The
373
combination of these factors underscores the importance of ensuring quality diabetes care extends to vulnerable populations, including those residing in rural areas. Promising practices in improving diabetes care have been noted; however, sustainability of quality initiatives can be elusive in the absence of organizational and technical support [14]. As national discussions regarding health care reform and quality of care continues, efforts to improve diabetes care among rural and underserved populations warrant consideration and should be supported. Acknowledgments The research reported here was supported in part by grant no U1CRH03711from the Office of Rural Health Policy, USDHHS.
References 1. Narayan, K. M., Boyle, J. P., Thompson, T. J., Sorensen, S. W., & Williamson, D. F. (2003). Lifetime risk for diabetes mellitus in the United States. Journal of the American Medical Association, 290(14), 1884–1890. 2. Johnson, E. A., Webb, W. L., McDowall, J. M., et al. (2005). A field-based approach to support improved diabetes care in rural states. Preventing Chronic Disease, 2(4), A08. 3. American Diabetes Association. (2008). Standards of Medical Care in Diabetes. Diabetes Care, 31(Suppl 1), S12–S54. 4. Lee, J. A., Liu, C. F., & Sales, A. E. (2006). Racial and ethnic differences in diabetes care and health care use and costs. Preventing Chronic Disease, 3(3), A85. 5. Mainous, A. G., 3rd, King, D. E., Garr, D. R., & Pearson, W. S. (2004). Race, rural residence, and control of diabetes and hypertension. Annals of Family Medicine, 2(6), 563–568. 6. Mukhtar, Q., Jack, L., Jr., Martin, M., Murphy, D., & Rivera, M. (2006). Evaluating progress toward Healthy People 2010 national diabetes objectives. Preventing Chronic Disease, 3(1), A11. 7. Phillips, M. M. (2004). Targeting diabetes preventive care programs: insights from the 2001 Behavioral Risk Factor Surveillance Survey. Preventing Chronic Disease, 1(1), A07. 8. Saaddine, J. B., Cadwell, B., Gregg, E. W., Engelgau, M. M., Vinicor, F., Imperatore, G., et al. (2006). Improvements in diabetes processes of care and intermediate outcomes: United States, 1988–2002. Annals of Internal Medicine, 144(7), 465–474. 9. Pathman, D. E., Ricketts, T. C., 3rd, & Konrad, T. R. (2006). How adults’ access to outpatient physician services relates to the local supply of primary care physicians in the rural southeast. Health Services Research, 41(1), 79–102. 10. Andrus, M. R., Kelley, K. W., Murphey, L. M., & Herndon, K. C. (2004). A comparison of diabetes care in rural and urban medical clinics in Alabama. Journal of Community Health, 29(1), 29–44. 11. Bell, R. A., Camacho, F., Goonan, K., Duren-Winfield, V., Anderson, R. T., Konen, J. C., et al. (2001). Quality of diabetes care among low-income patients in North Carolina. American Journal of Preventive Medicine, 21(2), 124–131. 12. Chin, M. H., Auerbach, S. B., Cook, S., et al. (2000). Quality of diabetes care in community health centers. American Journal of Public Health, 90(3), 431–434. 13. Coon, P., & Zulkowski, K. (2002). Adherence to American Diabetes Association standards of care by rural health care providers. Diabetes Care, 25(12), 2224–2229. 14. Kirkman, M. S., Williams, S. R., Caffrey, H. H., & Marrero, D. G. (2002). Impact of a program to improve adherence to diabetes
123
374
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
J Community Health (2010) 35:365–374 guidelines by primary care physicians. Diabetes Care, 25(11), 1946–1951. Bodenheimer, T., Wagner, E. H., & Grumbach, K. (2002). Improving primary care for patients with chronic illness: The chronic care model, part 2. Journal of the American Medical Association, 288(15), 1909–1914. Norris, S. L., Engelgau, M. M., & Narayan, K. M. (2001). Effectiveness of self-management training in type 2 diabetes: A systematic review of randomized controlled trials. Diabetes Care, 24(3), 561–587. Strine, T. W., Okoro, C. A., Chapman, D. P., Beckles, G. L., Balluz, L., & Mokdad, A. H. (2005). The impact of formal diabetes education on the preventive health practices and behaviors of persons with type 2 diabetes. Preventive Medicine, 41(1), 79–84. Balamurugan, A., Rivera, M., Jack, L., Jr., Allen, K., & Morris, S. (2006). Barriers to diabetes self-management education programs in underserved rural Arkansas: Implications for program evaluation. Preventing Chronic Disease, 3(1), A15. Butcher, M. K., Gilman, J., Meszaros, J. F., Bjorsness, D., Madison, M., McDowall, J. M., et al. (2006). Improving access to quality diabetes education in a rural state: The Montana Quality Diabetes Education Initiative. The Diabetes Educator, 32(6), 963–967. Powell, M. P., Glover, S. H., Probst, J. C., & Laditka, S. B. (2005). Barriers associated with the delivery of Medicare-reimbursed diabetes self-management education. The Diabetes Educator, 31(6), 890–899. Economic Research Service. (2008). Urban Influence Codes. U.S. Department of Agriculture. At http://www.ers.usda.gov/Briefing/ Rurality/UrbanInf. Accessed: April 2008. Economic Research Service. (2008). Measuring rurality: New definitions in 2003. U.S. Department of Agriculture. At: http://www.ers. usda.gov/Briefing/Rurality/NewDefinitions. Accessed: April 2008. Anderson, R. (1995). Revisiting the behavioral model and access to medical care: Does it matter? Journal of Health and Social Behavior, 36, 1–10. US Health Resources and Services Administration. (1993). Criteria for designation of areas having shortages of primary medical care professionals. In Relevant excerpts from 42 code of federal regulations (CFR), Chap. 1, Part 5, Appendix A (pp. 34–48). Mainous, A. G., 3rd, Koopman, R. J., Gill, J. M., Baker, R., & Pearson, W. S. (2004). Relationship between continuity of care
123
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
and diabetes control: Evidence from the Third National Health and Nutrition Examination Survey. American Journal of Public Health, 94(1), 66–70. Probst, J. C., Laditka, S. B., Wang, J. Y., & Johnson, A. O. (2007). Effects of residence and race on burden of travel for care: Cross sectional analysis of the 2001 US National Household Travel Survey. BMC Health Services Research, 7, 40. Weingarten, J. P., Jr., Brittman, S., Hu, W., Przybyszewski, C., Hammond, J. M., & FitzGerald, D. (2006). The state of diabetes care provided to Medicare beneficiaries living in rural America. Journal of Rural Health, 22(4), 351–358. US Department of Health and Human Services. (2000). Understanding and improving health and objectives for improving health. In Healthy people 2010 (2nd ed.). Washington, DC: U.S. Government Printing Office. Sloan, F. A., Grossman, D. S., & Lee, P. P. (2009). Effects of receipt of guideline-recommended care on onset of diabetic retinopathy and its progression. Ophthalmology, 116(8), 1515–1521, e1511–1513. Matza, L. S., Rousculp, M. D., Malley, K., Boye, K. S., & Oglesby, A. (2008). The longitudinal link between visual acuity and health-related quality of life in patients with diabetic retinopathy. Health and Quality of Life Outcomes, 6, 95. Probst, J. C., Moore, C. G., Baxley, E. G., & Lammie, J. J. (2002). Rural-urban differences in visits to primary care physicians. Family Medicine, 34(8), 609–615. Kirkbride, K., & Wallace, N. (2009). Rural health clinics and diabetes-related primary care for medicaid beneficiaries in Oregon. Journal of Rural Health, 25(3), 247–252. Chin, M. H., Cook, S., Drum, M. L., Jin, L., et al. (2004). Improving diabetes care in midwest community health centers with the health disparities collaborative. Diabetes Care, 27(4), 2–8. Siminerio, L. M., Piatt, G., & Zgibor, J. C. (2005). Implementing the chronic care model for improvements in diabetes care and education in a rural primary care practice. The Diabetes Educator, 31(2), 225–234. Balamurugan, A., Hall-Barrow, J., Blevins, M. A., et al. (2009). A pilot study of diabetes education via telemedicine in a rural underserved community—opportunities and challenges: A continuous quality improvement process. The Diabetes Educator, 35(1), 147–154.
