Demographic Results
The project population includes nearly 440,000 children and adults who were IHS active users during FY2010 and lived in one of the 14 project sites. Approximately one-third of the population was comprised of children and adolescents. See Table 1. The percentage of the population aged 65 years and older was 6.9 percent. Over half of the project population was female.
The age distribution of the project population at each project site varied. See Table 2. The percentage of children and adolescents ranged from a low of 28.0 percent to a high of 39.7 percent. Similarly, the percentage of the population aged 65 years and older also varied. At five project sites, this percentage was under 6 percent; at four other sites, the percentage was higher than 8 percent.
Since children and adolescents represented a large percentage of the project population, we assessed the percentage of adults who were aged 65 and older among all adults (i.e., aged 18 years and older). The percentage of adults aged 65 years and older among all adults averaged 10.2 percent across all project sites, and the percentage varied by project site.
Health Coverage
Health coverage was assessed for persons who used health services in FY2010. Health coverage is an indicator of the potential for a facility to obtain reimbursement for provided services, and the ability of a patient to have coverage for non-IHS services. Table 3 includes findings on health coverage for all project sites; Table 4 provides site-specific findings. Nearly 30 percent of those who used services in FY2010 had at least one day of Medicaid coverage during the year, and 10.1 percent had Medicare coverage. Nearly 15 percent had at least one day of private insurance coverage during FY2010. Just over half had no recorded type of coverage.
As can be seen in Table 3, Medicare and Medicaid coverage varied by age due to program eligibility. Among children and adolescents, 44.2 percent had Medicaid coverage. Among females aged 18 to 34 years, 38.5 percent had Medicaid coverage, while the average for both genders in that age group was 31.7 percent. Among persons aged 65 years and older, 93.5 percent had Medicare coverage.
Rates of health coverage varied by project site. The variation may be due to location, employment histories and opportunities, and health coverage enrollment processes at the facilities. Medicaid coverage among children and adolescents ranged from 20.4 percent to nearly 65 percent. Medicare coverage among adults aged 65 years and older was lower than 90 percent at only two sites, and was over 95 percent at two sites. The percentage of patients with no coverage ranged 29.8 percent to over 60 percent.
Health Status
We created three types of measures to assess the health status of the project population. The first was based on diagnosed conditions, the second on health risk, or morbidity burden, and the third on clinical outcome measures for glycemic, blood pressure, and cholesterol control.
The prevalence of diabetes among the AI/AN adults at the 14 project sites was 14.7 percent in FY2010.a See Table 5. Approximately 20 percent of adults aged 45-54 years were found to have diabetes; the prevalence was 32.9 percent among those aged 55-64 years and 37.5 percent among those aged 65 years and older. Ten percent of adults were found to have CVD; this number included persons with and without diabetes. Among adults aged 55-64 years, 11.5 percent had both diabetes and CVD. Approximately 20 percent of adults aged 65 years and older had both conditions. Of all adults, 5.4 percent had CVD but not diabetes.
Table 6 includes findings on the prevalence of diabetes and CVD among adults by project site. The prevalence of diabetes among adults ranged from a low of 6.6 percent at one site to over 20 percent at three project sites. The prevalence of CVD and the percentage of adults with both conditions also varied.
Information on the prevalence of selected health conditions, or comorbidities, among adults with or without diabetes is provided in Table 7. Nearly 80 percent of adults with diabetes had hypertension, and 31.6 percent had CVD. Over 13 percent had either renal disease or renal failure; the prevalence of renal failure was 3.3 percent. The prevalence of amputations is based on diagnoses related to lower limb amputations that occurred during a fiscal year and treatment of complications from previous lower limb amputations; it was found to be 2.4 percent among adults with diabetes.
The prevalence of mental health and substance use disorders among adults with diabetes was 23.5 percent and 5.6 percent, respectively. Nearly 16 percent of adults with diabetes were diagnosed with depression, one of the identified mental health disorders noted above. The prevalence of liver disease was 6.8 percent. Site-specific information on the prevalence of selected conditions among adults with diabetes is provided in Table 8.
Each person was assigned a health risk score to reflect his or her morbidity burden and expected use of health resources. The health risk score is benchmarked to a U.S. commercially insured population, which had an average health risk of 1.0. Patients with more high-cost acute and chronic conditions have higher risk than those without such conditions. The average health risk of the FY2010 AI/AN project population was 1.3, or 30 percent higher than the average risk for the U.S. commercially insured population. See Table 9. The average health risk of AI/ANs with diabetes was 4.6, or more than four times higher than that of the reference population. The average health risk of those with CVD but not diabetes was 5.0.
Table 10 includes data on health risk by project site. The health risk of persons with diabetes ranged from 3.4 to 7.1 across the project sites. For three project sites, the health risk among those with diabetes was 6 or higher. Although the average risk among those with both diabetes and CVD averaged 8.8, in eight project sites the morbidity burden was 10 or higher.
It is important to note that a number of factors, other than the actual prevalence of disease, influenced project site variations in the prevalence of diabetes, CVD, other conditions, and health risk. One factor is the quality of diagnostic code data (e.g., some project sites may have more resources allocated to documenting diagnostic codes). A second is utilization of non-IHS services. The project data include diagnostic codes for I/T and CHS services and do not include diagnostic codes for services obtained from non-IHS providers that are not paid for by CHS.
The health status of persons with diabetes was also assessed using data on clinical measures for glycemic, blood pressure, and cholesterol control. Findings concerning these three measures for adults with diabetes are presented in Tables 11 and 12. Data availability for these measures varied across project sites and data for four sites were excluded from both tables.
Although optimal levels for glycemic control (HbA1c), blood pressure, and cholesterol are determined by a patient’s health status, age, and other factors, IHS general treatment guidelines were used to report on these clinical measures for adults with diabetes. Nearly 80 percent of adults with diabetes had an HbA1c test result. Of these, 63.3 percent had HbA1c values less than 8.0 percent. Of the 97.2 percent of adults with diabetes with blood pressure results, nearly 71 percent had a systolic blood pressure less than 140 mmHg, and of the 64.5 percent of adults with diabetes with a cholesterol result, 59.5 percent were found to have LDL cholesterol values less than 100 mg/dl. See Table 11.
Due to the importance of glycemic, blood pressure, and cholesterol control for maintaining health and preventing the onset of comorbidities, values for HbA1c, blood pressure, and LDL cholesterol were combined for adults with data on all three clinical measures in Table 12. Using the defined optimal values (i.e., less than 8.0 percent for HbA1c, less than 140 mmHg for systolic blood pressure, less than 100 mg/dl for LDL cholesterol), 31.1 percent of adults with diabetes had values within these ranges for all three measures. Forty percent met two of the three guidelines and 23.8 percent met only one. A small percentage of persons (i.e., 5.0 percent) had no values within the optimal ranges.
Similar to other findings, there was significant variation across the project sites in measures of HbA1c, blood pressure, and cholesterol. It is important to note that Table 11 and Table 12 findings may vary from those reported in the IHS Diabetes Care and Outcomes Audit by the p. There are two main reasons for the variation. First, the population included in the Audit differs from the population included in this project (i.e., active users who lived in an IHS Service Unit during a specific fiscal year). Second, the Audit is based on data compiled with Service Unit data rather than NDW data. Not all Service Unit blood pressure and laboratory test values are recorded in data fields provided to the NDW. Consequently, the AUDIT report may include data not available from the NDW.
Information on health status was combined with that on health coverage to ascertain the prevalence of health coverage among adults with diabetes or CVD. See Table 13. Sixty percent of adults with diabetes or CVD had some form of health coverage. As expected, the percentage was high for persons aged 65 years and older due to Medicare coverage. Less than 50 percent of adults aged 18 to 64 years with diabetes or CVD had health coverage.
Health Service Utilization
NDW and CHS data were used to assess health service utilization among all persons, both children and adults, and among those with diabetes and CVD. Tables 14-18 include information on inpatient service utilization; findings concerning utilization of outpatient services and prescribed medications are provided in Tables 19-21. While some information on ECM services is provided in Table 19, Chapter 6 includes detailed ECM findings. It is important to note that service utilization for persons who lived in the 14 project sites includes services they obtained in that project site and in other IHS Service Units, even if the other Service Units were not among the 14 project sites.
Utilization of Hospital Inpatient Services
Table 14 provides an overview of inpatient service utilization based on data for admissions to I/T facilities and admissions paid for by CHS. Among persons with diabetes, the admission rate for I/T inpatient services was 0.13 admissions per person, and the admission rate for CHS inpatient services was 0.10 admissions per person. Thus, the hospital inpatient admission rate for persons with diabetes was 0.23 admissions per person when I/T admissions and CHS admissions were combined. Similarly, the average number of I/T inpatient days in FY2010 was 0.6 and the average number of CHS inpatient days was 0.4. Thus, on average, persons with diabetes spent one day in the hospital during FY2010 (i.e., 0.6+0.4=1.0). The average number of inpatient days for persons with diabetes and CVD was 2.3, more than twice that of persons with diabetes, regardless of CVD status. Persons with CVD but not diabetes averaged 1.3 days in the hospital during FY2010.
Information on utilization of I/T inpatient services by project site is provided in Table 15. As noted above, the types of inpatient services provided across service units vary. Some hospitals provide a more comprehensive array of inpatient services including obstetrical care, while other facilities may provide more limited services. On average, the percentage of hospital admissions that were readmissions within 30 days of a previous admission was 9.6 percent. Information on the type of inpatient stay was used to identify pediatric, obstetric, and adult non-obstetric admissions. Two-thirds of inpatient stays were for adults who were admitted for reasons other than obstetrical care.
Tables 16-17 provide additional information on I/T adult non-obstetrical inpatient admissions. According to the Table 16 findings, over 60 percent of adult admissions for non-obstetric care were by adults with diabetes or CVD. The percentage varied by project site. Some of the variation across project sites is due to differences in the prevalence of diabetes and CVD; some variation may be due to differences in practice patterns and access to and use of outpatient services.
Using a nationally recognized algorithm developed by AHRQ, we categorized adult non-obstetric admissions as admissions that were and were not sensitive to ambulatory care. According to the AHRQ algorithm, 23.4 percent of all adult non-obstetric admissions were classified as sensitive to ambulatory care services. See Table 17. The percentage among adults with diabetes or CVD was higher, with 30.8 percent of admissions classified as sensitive to ambulatory services. This finding is particularly noteworthy in that admissions by adults with diabetes or CVD represent over 60 percent of all adult non-obstetric inpatient admissions. Table 18 provides information on ambulatory-sensitive hospital admissions by project site. Among adults with diabetes or CVD, the percentage of non-obstetric admissions classified as ambulatory sensitive ranged from a low of 21.8 percent to over 60 percent across the 12 project sites.
Utilization of Outpatient Services and Prescribed Medications
Table 19 provides information on the number of visits and average utilization rates for FY2010 outpatient services for all 14 project sites. Detailed utilization data are provided for I/T services while CHS outpatient service utilization is summarized.j As this report focuses on services for AI/ANs with diabetes or CVD, outpatient utilization by persons with these conditions is highlighted.
On average, persons with diabetes had 0.9 ED visits and 0.5 urgent care visits. When both types of services are added together, persons with diabetes had, on average, nearly 1.5 visits annually for ED and urgent care services. Persons with CVD but not diabetes had similar use of ED and urgent care services.
Among persons with diabetes, the average number of primary care clinic visits, including general office and diabetes clinic visits, was 5.8. Encounters in clinics classified as primary or general services averaged 4.8 for persons with diabetes. Although not all project sites offered services through a diabetes clinic,k use of diabetes clinic services averaged 1.0 visit annually across project sites. The average number of primary care clinic visits among those with both diabetes and CVD was somewhat higher (i.e., 7.1). Among those with CVD but not diabetes, the average number was 5.5.
AI/ANs with diabetes averaged 0.9, 0.9, and 0.2 visits to dental, eye, and podiatry or diabetes foot clinics, respectively, during FY2010. They had, on average, 1.2 ECM visits. More information on ECM utilization is provided in Chapter 6.
Utilization of behavioral health, physical therapy, and other rehabilitative services varied somewhat by health status, although the variation among those with diabetes or CVD was not as great as that for other services. For example, the average number of behavioral health visits for persons with diabetes during FY2010 was 0.3. Persons with CVD but not diabetes also had, on average, 0.3 behavioral health clinic visits. The average number of physical therapy visits by persons with diabetes was the same as that for persons with CVD but not diabetes (i.e., 0.3 visits). The average number of home visits among those with diabetes was 0.4. However, those with diabetes and CVD averaged 0.8 home visits during FY2010.
As expected, persons with diabetes or CVD had higher use of prescribed medications than those without these conditions. The number of prescriptionsl dispensed by I/T providers for persons with diabetes averaged 57.4 during FY2010. The average number for persons with CVD but not diabetes was 40.4. More detailed information on prescribed medications is provided in Tables 20 and 21. The number of dispensed prescribed medications by the Veterans Administration Therapeutic Medication Classifications is provided in Table 20. Over half of dispensed medications were classified as 1) cardiovascular (21.0 percent; e.g., ACE inhibitors, antilipemic agents, beta blockers); 2) central nervous system (19.0 percent; e.g., antidepressants, non-opioid analgesics); or 3) hormones, synthetics, or modifiers (11.3 percent; e.g., insulin, oral hypoglycemic agents).
Information on specific types of medications dispensed for adults with diabetes is provided in Table 21 by project site. Among adults with diabetes, 79.8 percent were dispensed diabetes-related medications, 76.9 percent were dispensed blood pressure-related medications, and 56.1 percent were dispensed cholesterol-related medications. Similar to other utilization indicators, variations across the project sites were found with regard to prescribed medications for adults with diabetes.
IHS Total Treatment Costs
Annual IHS total treatment costs, including I/T and CHS costs, were estimated for all persons and for adults by health status. FY2010 IHS total treatment costs for adults with diabetes were estimated to be $8,164. The costs for adults without diabetes were estimated to be $2,789. See Table 22. The 14.7 percent of adults with diabetes accounted for one-third of all treatment costs for adults. The costs of treating an adult with diabetes and CVD were approximately twice those of treating an adult with diabetes but not CVD; the estimated costs were $12,568 and $6,129, respectively. Among adults with CVD but not diabetes, IHS total treatment costs averaged $9,009.
We examined how IHS total treatment costs varied by the presence of specific comorbidities in two ways. First, we reported costs for specific combinations of chronic disease; the chronic condition combinations did not identify exclusive groups of patients. As shown in Table 22, costs for adults with diabetes and a mental health disorder averaged $11,120. Costs for adults with diabetes and a substance use disorder were estimated to be $14,216. Those for adults with diabetes and renal disease or renal failure, excluding the costs of dialysis, averaged $14,339. It is important to note that adults with diabetes who had these comorbidities may also have had other comorbidities (e.g., hypertension, CVD). For this reason, we also examined IHS total treatment costs for adults with diabetes by disease stage. Eight stages were defined based on the presence of hypertension, diabetes, CVD, amputation-related procedures, and renal disease or renal failure. The eight stages identified exclusive patient groups. See Appendix B. The estimated costs of treating an adult with diabetes and hypertension, but not with CVD, amputation-related procedures, renal disease, or renal failure, were $6,288. The costs of treating adults with diabetes and CVD, but not with any amputation-related procedures, renal disease, or renal failure, regardless of their hypertension status, were estimated to be $10,391. The cost estimates for those with diabetes and amputation-related procedures, renal disease, or renal failure were greater.
Table 22 also includes findings on health risk. The health risk score provides information on morbidity burden based on the presence of acute and chronic conditions, and expected use of health resources based on the costs of treating the commercially insured population. As can be seen in Table 22, the health risk among adults with diabetes was 3.7 times that of adults without diabetes (i.e., 4.6 divided by 1.3 is 3.7). Average estimated IHS total treatment costs for adults with diabetes were 2.9 times those of adults without the condition (i.e., $8,164 divided by $2,789 is 2.9).
aThis is the prevalence among adult active users. The prevalence among adults who used IHS services in FY2010 was 17.9 percent.
j Unless otherwise noted, outpatient services described by type of service are those for I/T providers.
k Some Service Units that implemented the IPC model of care no longer sponsor diabetes clinic and now provide such services during primary care visits
l Prescriptions include those for medications and supplies.
