Abstract
Background:
The burden of diabetes is exceptionally high among American Indian and Alaska Native peoples (AI/ANs). The Indian Health Service (IHS) and Tribal health programs provide education, case management, and advanced practice pharmacy (ECP) services for AI/ANs with diabetes to improve their health outcomes.
Objective:
Evaluate patient outcomes associated with ECP use by AI/AN adults with diabetes.
Research Design:
This observational study included the analysis of IHS data for fiscal years (FY) 2011–2013. Using propensity score models, we assessed FY2013 patient outcomes associated with FY2012 ECP use, controlling for FY2011 baseline characteristics.
Subjects:
AI/AN adults with diabetes who used IHS and Tribal health services (n=28,578).
Measures:
We compared health status and hospital utilization outcomes for ECP users and non-users.
Results:
Among adults with diabetes, ECP users, compared to non-users, had lower odds of high systolic blood pressure (OR=0.85, p<0.001) and high low-density lipoprotein cholesterol (OR=0.89, p<0.01). Among adults with diabetes absent cardiovascular disease (CVD) at baseline, 3 or more ECP visits, compared to no visits, was associated with lower odds of CVD onset (OR=0.79, p<0.05). Among adults with diabetes and CVD, any ECP use was associated with lower odds of end-stage renal disease onset (OR=0.60, p<0.05). ECP users had lower odds of 1 or more hospitalizations (OR=0.80, p<0.001).
Conclusions:
Findings on positive patient outcomes associated with ECP use by adults with diabetes may inform IHS and Tribal policies, funding, and enhancements to ECP services to reduce disparities between AI/ANs and other populations in diabetes-related morbidity and mortality.
Keywords: American Indian and Alaska Native, diabetes, patient education, cardiovascular disease, observational studies
Introduction
American Indian and Alaska Native (AI/AN) peoples experience some of the greatest health disparities with respect to diabetes and related complications.1–5 The prevalence of diabetes among AI/ANs aged 18 years and older was 14.7% in 2017–2018, nearly double that of non-Hispanic whites and the highest among U.S. racial/ethnic groups.1 The AI/AN all-cause mortality rate is 46% higher than that of non-Hispanic whites and is largely attributable to disparities in heart disease, stroke, diabetes, and kidney disease mortality.2,6–8 Additionally, diabetes and cardiovascular disease (CVD) contribute to higher rates of premature mortality.6–8
Many AI/ANs obtain health care through services funded by the Indian Health Service (IHS). The IHS service delivery system includes hospitals, clinics, and health programs operated by the federal government, Tribal organizations, and urban Indian health programs. Known collectively as I/T/Us, they serve approximately 2.6 million AI/ANs throughout the United States.9 IHS and Tribes support an array of services to implement IHS Standards of Care for diabetes to reduce complications among those with diabetes.10 The Special Diabetes Program for Indians (SDPI) provides grants to over 300 I/T/Us and community-based programs to support diabetes prevention and treatment, including education, case management, and advanced practice pharmacy (ECP) services.5 Since SDPI’s implementation in 1998, intermediate clinical outcomes (e.g., blood glucose and cholesterol levels) among AI/AN adults with diabetes have improved;11 hospitalizations for uncontrolled diabetes have declined;12 and the incidence of diabetes-related end-stage renal disease (ESRD) has substantially decreased, as have estimated ESRD-related Medicare expenditures.13
In non-AI/AN populations, ECP services have improved outcomes among individuals with diabetes.14–21 Information about the provision and use of ECP services among AI/ANs with diabetes who access I/T/U services is increasingly available.11,22–26 A 2014 study of the SDPI Healthy Heart demonstration project linked intensive case management with improvements in CVD risk factors among approximately 3,400 adults with diabetes.23 Participants had reductions in high blood sugar, blood pressure, and cholesterol after 12 months of case management services. Though the study documented a successful translation of an intensive case management program, it did not include a comparison population to provide context for the findings. Thus, information on patient outcomes associated with ECP utilization by adults with diabetes for a larger, more representative sample is needed to guide enhancements to ECP services and effectively allocate I/T/U resources.
This is particularly important as AI/ANs with diabetes who use I/T/U services require delivery models that can effectively address their risks and complex needs. IHS resources are strained due to limited per capita spending ($4,078 in fiscal year [FY] 2017).9,25 Although this amount does not include all spending associated with patient care, it is substantially lower than per capita spending for the U.S. general population ($10,742) in 2017.27 IHS resources are further compromised by provider shortages and community-level factors that affect patient service use and health (e.g., low household income, rural geography).25,28–33
To add to this emerging literature, we evaluated patient outcomes associated with ECP utilization by analyzing IHS data for a large, geographically diverse group of AI/AN adults with diabetes (n=28,578). Using an observational design, we compared the health status and hospital utilization of ECP users to non-users to assess outcomes associated with using ECP services in addition to usual care (e.g., primary care and specialty services).
Methods
This study was approved by the IHS National Institutional Review Board (IRB), Tribal IRBs, and Tribal Councils and Authorities, in addition to the university’s IRB.
A. Data
This study was conducted using data extracted from a longitudinal data infrastructure that houses health status, service utilization, and treatment cost data for over 640,000 AI/ANs who live throughout the United States, representing nearly 30% of AI/ANs who use IHS services.24 The data infrastructure, created as the IHS Improving Health Care Delivery Data Project, is a synthesis of existing health data from multiple IHS platforms and includes data for FY2007-FY2013.
The data infrastructure includes information for a purposeful sample of AI/ANs who lived in 15 IHS Service Units. IHS Service Units, which are health service administrative units defined by geographic areas that include one or more health facilities, are located throughout the United States. One Service Unit is located in the East, 4 in the Northern Plains, 2 in the Southern Plains, 5 in the Southwest, 2 in the Pacific Coast, and 1 in Alaska.2 The IHS Data Project population was identified by geographic area, rather than by random sampling, to create important community-level (e.g., drive time to services) and county-level (e.g., household income) measures not available elsewhere. Communities are defined geographic areas within Service Units (hereafter we refer to Service Units as project sites). Sources of IHS electronic data include the National Data Warehouse (NDW) for data on registration and services rendered by I/T providers and the Purchased/Referred Care (PRC) program for data on non-I/T services paid for by some IHS and Tribal health programs; PRC program data for other projects sites was obtained from the PRC fiscal intermediary. In this study, we refer to I/T services, rather than I/T/U services, since the urban Indian clinics located in the 12 project sites included in the analysis provided very few ECP visits in FY2012. The IHS Data Project population is comparable to the national IHS service population in terms of age and sex.34
B. Study population
The study population included adults who 1) had diabetes in FY2011, 2) used I/T services during 3 consecutive fiscal years (FY2011-FY2013), and 3) lived in 1 of 12 project sites that provided ECP services and had complete data. Study population exclusion criteria included 1) treatment for malignant cancer, a transplant, or ESRD (except for the analysis of onset of ESRD in FY2013) during FY2011-FY2013; 2) having missing data for community- and county-level variables; and 3) evidence of having died during FY2013.
C. Measures
Data for all measures were extracted from the IHS Data Project data infrastructure for FY2011-FY2013, except where noted.
Demographic and health status.
NDW data provided information on age, gender, and health insurance coverage in FY2011. Project specific algorithms, developed from national references, were used to identify adults with diabetes, CVD, and ESRD using ICD-9-CM diagnoses, procedure codes, medication use, and blood glucose control included in the NDW and PRC service utilization records.35–37 The diabetes and CVD measures were used to create 3 study cohorts: all adults with diabetes and 2 subgroups‒adults with diabetes absent CVD and adults with both diabetes and CVD.
SightlinesTM DxCG Risk Solutions software38 was used to identify patients diagnosed with other conditions (e.g., hypertension, renal disease). The DxCG software also provided a measure of morbidity burden (i.e., health risk score) for individuals based on their age, gender, and all diagnosed conditions. For each health status group, we categorized the risk scores into quartiles; adults with the lowest morbidity were assigned to quartile 1 and those with the highest morbidity to quartile 4.
Patient outcome measures included 5 diabetes management indicators. Based on IHS and national guidelines for diabetes management, diabetes management indicators include high hemoglobin A1c (≥ 8%), high systolic blood pressure (SBP, ≥ 140 mmHg), and high low-density lipoprotein (LDL) cholesterol (≥ 100 mg/dL).10,11,39 New or recurring onset of CVD was defined as having CVD during FY2013 with no diagnosis of the condition during the previous 3 fiscal years. ESRD onset in FY2013 was assessed using data for all previous years (FY2007–2013).
Health service utilization.
ECP utilization was defined using I/T data on visits for individual or group diabetes education, provided by nurses or health educators in diabetes clinics; nutrition education; advanced practice pharmacy (APP); case management; and other types of education (e.g., smoking cessation, obesity). APP delivery models varied and could have included patient assessment, medication reconciliation, and health education conducted by certified pharmacists who may also have ordered laboratory tests and modified prescriptions under the supervision of a physician or as primary care providers.26,31,40,41 Thus, ECP visits do not include education provided as part of other outpatient visits with physicians and mid-level providers (e.g., primary and specialty care visits). We created 2 ECP utilization measures: any ECP use (a dichotomous measure) and level of ECP use (no visits, 1–2 visits, 3+ visits).
NDW I/T and PRC non-I/T inpatient data were combined to create 2 hospital inpatient utilization measures: 1 or more hospitalizations and total number of hospital inpatient days. I/T hospital data were used to create 2 additional measures: 1 or more potentially preventable hospitalizations24 and number of emergency department (ED) visits.
Facility, community, and county measures.
We created 2 measures of ECP access in FY2012. The facility ECP supply rate was calculated as the number of provided ECP visits divided by the number of adults living in the facility’s service area, defined by communities. Patient drive time to ECP services was estimated from a central location in each community to an I/T facility that provided ECP using geocodes (latitude and longitude).42
County-level measures of AI/AN educational attainment and household income were derived from 2010–2014 American Community Survey county-level data for AI/ANs who reported access to IHS services.43 The educational attainment measure is the percentage of adults aged 25 years and older who did not complete high school. We defined the percentage of households with low income as the percentage with incomes below 139% of the federal poverty level, a poverty level used in many states to determine one type of Medicaid eligibility.
D. Analysis
We used SAS® and Stata statistical software to conduct descriptive and multivariate analyses.44,45 This study employed an observational design to compare diabetes management and hospital utilization outcomes among ECP users and non-users. To address nonrandom assignment of patients to 2 groups based on FY2012 ECP use (i.e., ECP users and non-users), we used inverse probability of treatment weighted (IPTW) estimation.46,47 IPTW is a statistical approach often used when considering causal effects in observational studies without randomization whereby the distributions of potential confounders between comparison groups are statistically balanced.47
Equation 1 of the model estimated a patient’s propensity to use ECP services during FY2012, based on baseline (i.e., FY2011) patient and provider characteristics. Equation 2 compared patient outcomes during FY2013 between FY2012 ECP users and non-users, after adjusting for the baseline differences between the comparison groups using IPTW estimation. To address residual confounding, Equation 2 also included baseline patient characteristics. The propensity models included fixed effects in both equations to control for variations across project sites.
We estimated 2 sets of propensity models. The first set examined patient outcomes associated with any ECP use, compared to no use, using logistic regression for Equation 1. The second set estimated relationships between level of ECP use and patient outcomes using ordered logistic regression for Equation 1. For both sets, the specification of Equation 2 varied by patient outcome. Binary outcomes were estimated using logistic regression; the number of ED visits and hospital inpatient days were modeled using negative binomial regression.
Results
Of all adults with diabetes, 28,578 adults met study inclusion and exclusion criteria. Additional information on the study sample is provided in the appendix. Statistical differences (p<0.001) were observed between ECP users and non-users in FY2011. Throughout this section, we first present results for all adults with diabetes. Next, we provide results for the subsamples absent CVD and with CVD, where noteworthy.
ECP users were older; had a higher prevalence of hypertension, CVD, and renal disease; and had a higher morbidity burden (Table 1). ECP users, compared to non-users, lived in communities where the mean drive time to a facility with ECP services was shorter (15.2 and 22.5 minutes, p<0.001), and in counties where a smaller percentage of households had low household incomes (41.1% and 45.1%, p<0.001).
Table 1.
Baseline (fiscal year 2011, except where noted) characteristics of adults with diabetes by cardiovascular disease (CVD) status and use of education, case management, and advanced practice pharmacy (ECP) services during fiscal year 2012.
| Adults with diabetes | Adults with diabetes absent CVD | Adults with diabetes and CVD | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Characteristic | ECP non-users | ECP users | ECP non-users | ECP users | ECP non-users | ECP users | |||||||||
| N | % | N | % | N | % | N | % | N | % | N | % | ||||
| All adults with diabetes | 16,855 | 100.0% | 11,723 | 100.0% | 12,733 | 100.0% | 8,044 | 100.0% | 4,122 | 100.0% | 3,679 | 100.0% | |||
| Gender | * | ||||||||||||||
| Female | 9,635 | 57.2% | 6,795 | 58.0% | 7,506 | 59.0% | 4,861 | 60.4% | 2,129 | 51.7% | 1,934 | 52.6% | |||
| Male | 7,220 | 42.8% | 4,928 | 42.0% | 5,227 | 41.1% | 3,183 | 39.6% | 1,993 | 48.4% | 1,745 | 47.4% | |||
| Age | *** | *** | *** | ||||||||||||
| 18–34 years | 1,728 | 10.3% | 827 | 7.1% | 1,552 | 12.2% | 722 | 9.0% | 176 | 4.3% | 105 | 2.9% | |||
| 35–44 years | 2,876 | 17.1% | 1,615 | 13.8% | 2,502 | 19.7% | 1,345 | 16.7% | 374 | 9.1% | 270 | 7.3% | |||
| 45–54 years | 4,291 | 25.5% | 2,917 | 24.9% | 3,506 | 27.5% | 2,238 | 27.8% | 785 | 19.0% | 679 | 18.5% | |||
| 55–64 years | 4,245 | 25.2% | 3,355 | 28.6% | 3,080 | 24.2% | 2,244 | 27.9% | 1,165 | 28.3% | 1,111 | 30.2% | |||
| 65+ years | 3,715 | 22.0% | 3,009 | 25.7% | 2,093 | 16.4% | 1,495 | 18.6% | 1,622 | 39.4% | 1,514 | 41.2% | |||
| Health Coverage | |||||||||||||||
| No coverage in addition to IHS access | 6,457 | 38.3% | 4,162 | 35.5% | *** | 5,311 | 41.7% | 3,213 | 39.9% | 1,146 | 27.8% | 949 | 25.8% | * | |
| Medicaid | 3,684 | 21.9% | 2,107 | 18.0% | *** | 2,832 | 22.2% | 1,525 | 19.0% | *** | 852 | 20.7% | 582 | 15.8% | *** |
| Medicare | 4,805 | 28.5% | 4,016 | 34.3% | *** | 2,808 | 22.1% | 2,071 | 25.8% | *** | 1,997 | 48.5% | 1,945 | 52.9% | *** |
| Private | 3,959 | 23.5% | 2,968 | 25.3% | *** | 3,047 | 23.9% | 2,115 | 26.3% | *** | 912 | 22.1% | 853 | 23.2% | |
| Health Condition | |||||||||||||||
| Hypertension | 11,798 | 70.0% | 9,081 | 77.5% | *** | 8,419 | 66.1% | 5,926 | 73.7% | *** | 3,379 | 82.0% | 3,155 | 85.8% | *** |
| Cardiovascular disease | 4,122 | 24.5% | 3,679 | 31.4% | *** | ||||||||||
| Renal disease | 1,223 | 7.3% | 1,147 | 9.8% | *** | 344 | 2.7% | 309 | 3.8% | *** | 879 | 21.3% | 838 | 22.8% | |
| Neuropathy | 2,287 | 13.6% | 2,002 | 17.1% | *** | 1,419 | 11.1% | 1,082 | 13.5% | *** | 868 | 21.1% | 920 | 25.0% | *** |
| Amputationsa | 213 | 1.3% | 159 | 1.4% | 76 | 0.6% | 45 | 0.6% | 137 | 3.3% | 114 | 3.1% | |||
| Mental health disordersb | 3,510 | 20.8% | 2,880 | 24.6% | *** | 2,540 | 20.0% | 1,914 | 23.8% | *** | 970 | 23.5% | 966 | 26.3% | ** |
| Depression | 2,350 | 13.9% | 2,022 | 17.3% | *** | 1,708 | 13.4% | 1,322 | 16.4% | *** | 642 | 15.6% | 700 | 19.0% | *** |
| Alcohol and drug abuse | 1,101 | 6.5% | 571 | 4.9% | *** | 841 | 6.6% | 380 | 4.7% | *** | 260 | 6.3% | 191 | 5.2% | * |
| Alcohol abuse | 861 | 5.1% | 436 | 3.7% | *** | 658 | 5.2% | 302 | 3.8% | *** | 203 | 4.9% | 134 | 3.6% | ** |
| Drug abuse | 357 | 2.1% | 200 | 1.7% | * | 267 | 2.1% | 120 | 1.5% | ** | 90 | 2.2% | 80 | 2.2% | |
| Tobacco use disorders | 1,528 | 9.1% | 1,028 | 8.8% | 1,075 | 8.4% | 635 | 7.9% | 453 | 11.0% | 393 | 10.7% | |||
| Liver disease | 700 | 4.2% | 588 | 5.0% | *** | 505 | 4.0% | 383 | 4.8% | ** | 195 | 4.7% | 205 | 5.6% | |
| Eye disease | 2,581 | 15.3% | 2,364 | 20.2% | *** | 1,660 | 13.0% | 1,373 | 17.1% | *** | 921 | 22.3% | 991 | 26.9% | *** |
| Risk score quartiles | *** | *** | *** | ||||||||||||
| Quartile 1 (lowest risk) | 5,067 | 30.1% | 2,071 | 17.7% | 3,738 | 29.4% | 1,455 | 18.1% | 1,280 | 31.1% | 669 | 18.2% | |||
| Quartile 2 | 4,356 | 25.8% | 2,793 | 23.8% | 3,265 | 25.6% | 1,930 | 24.0% | 1,027 | 24.9% | 924 | 25.1% | |||
| Quartile 3 | 3,857 | 22.9% | 3,289 | 28.1% | 2,967 | 23.3% | 2,227 | 27.7% | 936 | 22.7% | 1,014 | 27.6% | |||
| Quartile 4 (highest risk) | 3,575 | 21.2% | 3,570 | 30.5% | 2,763 | 21.7% | 2,432 | 30.2% | 879 | 21.3% | 1,072 | 29.1% | |||
| Clinical measures | |||||||||||||||
| Systolic blood pressure | *** | *** | *** | ||||||||||||
| No SBP result | 566 | 4.5% | 174 | 2.4% | 501 | 5.3% | 154 | 3.0% | 65 | 2.2% | 20 | 0.9% | |||
| <140 mmHg | 8,999 | 72.2% | 5,578 | 75.4% | 6,957 | 73.4% | 3,971 | 77.0% | 2,042 | 68.6% | 1,607 | 71.8% | |||
| ≥140 mmHg | 2,893 | 23.2% | 1,642 | 22.2% | 2,022 | 21.3% | 1,030 | 20.0% | 871 | 29.3% | 612 | 27.3% | |||
| A1c | *** | *** | *** | ||||||||||||
| No A1c result | 2,178 | 17.5% | 611 | 8.3% | 1,751 | 18.5% | 488 | 9.5% | 427 | 14.3% | 123 | 5.5% | |||
| <8% | 5,878 | 47.2% | 3,722 | 50.3% | 4,365 | 46.0% | 2,513 | 48.8% | 1,513 | 50.8% | 1,209 | 54.0% | |||
| ≥8% | 4,402 | 35.3% | 3,061 | 41.4% | 3,364 | 35.5% | 2,154 | 41.8% | 1,038 | 34.9% | 907 | 40.5% | |||
| LDL cholesterol | *** | *** | *** | ||||||||||||
| No LDL result | 3,819 | 30.7% | 1,555 | 21.0% | 3,041 | 32.1% | 1,202 | 23.3% | 778 | 26.1% | 353 | 15.8% | |||
| <100 mg/dL | 4,730 | 38.0% | 3,524 | 47.7% | 3,342 | 35.3% | 2,244 | 43.5% | 1,388 | 46.6% | 1,280 | 57.2% | |||
| ≥100 mg/dL | 3,909 | 31.4% | 2,315 | 31.3% | 3,097 | 32.7% | 1,709 | 33.2% | 812 | 27.3% | 606 | 27.1% | |||
| Hospital service use | |||||||||||||||
| Average # of I/T ED visits | 0.79 | 0.78 | 0.69 | 0.62 | ** | 1.10 | 1.12 | ||||||||
| 1 or more hospitalizations (%) | 1,226 | 8.8% | 883 | 10.3% | ** | 583 | 5.5% | 340 | 5.7% | 643 | 19.9% | 543 | 21.2% | ||
| Average # of hospital inpatient days | 0.49 | 0.61 | ** | 0.23 | 0.25 | 1.32 | 1.46 | ||||||||
| 1 or more I/T potentially preventable hospitalizations (%) | 334 | 2.1% | 296 | 2.6% | ** | 135 | 1.1% | 102 | 1.3% | 199 | 5.0% | 194 | 5.5% | ||
| Mean ECP Drive time 2012 (min) | 22.5 | 15.2 | *** | 22.3 | 15.4 | *** | 23.0 | 14.6 | *** | ||||||
| ECP FY2012 facility rate | 0.30 | 0.35 | *** | 0.30 | 0.34 | *** | 0.31 | 0.36 | *** | ||||||
| County-level demographic information | |||||||||||||||
| Mean percent with < high school completion | 46.8% | 46.0% | 46.8% | 46.0% | 46.8% | 45.9% | |||||||||
| Mean percent with < 139% federal poverty level | 45.1% | 41.1% | *** | 45.2% | 41.4% | *** | 44.7% | 40.4% | *** | ||||||
IHS: Indian Health Service; ECP: education, case management and advanced practice pharmacy; I/T: Indian Health Service and Tribal
p < 0.05;
p < 0.01;
p < 0.001
The prevalence of amputations represents amputations noted on the utilization records during the specific fiscal years; providers may not have documented in the utilization record that a person had an amputation during previous years.
Depression is one of the mental health disorders included in the category Mental health disorders; other types of mental health disorders include anxiety, bipolar, and post-traumatic stress disorders. Persons in the Depression category may or may not have other mental health disorders.
Adults with diabetes and CVD, compared to adults with diabetes absent CVD, were older and had higher rates of comorbidities. Despite this, many differences observed between ECP users and non-users among all adults with diabetes in FY2011 were also observed among adults with and without CVD.
During FY2012, 41.0% of adults with diabetes used ECP services (Table 2). Their average number of ECP visits was 2.8, with 68.6% having 1–2 visits, and 31.4% having 3 or more.
Table 2.
Use of education, case management, and advanced practice pharmacy (ECP) services among adults with diabetes. Fiscal year 2012.
| Adults with diabetes | Adults with diabetes absent CVD | Adults with diabetes and CVD | ||
|---|---|---|---|---|
| Average number of ECP visits | 1.1 | 0.9 | 1.7 | *** |
| Percent who had at least 1 ECP visit | 41.0% | 38.7% | 47.2% | *** |
| Average number of ECP visits among those who had at least 1 visit | 2.8 | 2.3 | 3.7 | *** |
| Distribution of the number of visits among those with at least one visit | *** | |||
| Percent who had 1 ECP visit | 45.4% | 48.3% | 38.9% | |
| Percent who had 2 ECP visits | 23.2% | 23.9% | 21.6% | |
| Percent who had 3 ECP visits | 11.1% | 11.1% | 11.3% | |
| Percent who had 4 or more ECP visits | 20.3% | 16.7% | 28.2% | |
| All | 100% | 100% | 100% |
CVD: cardiovascular disease
p < 0.001, statistical significance of differences between adults with and without CVD.
We compared FY2011 and FY2013 data for the 5 diabetes management and 4 hospital utilization outcome measures (Table 3). Between FY2011 and FY2013, the percentage of adults with diabetes with high SBP and high A1c increased. In contrast, the percentage with high LDL cholesterol decreased. Among adults with diabetes, a statistically lower percentage of ECP users, compared to non-users, had high SBP and high LDL cholesterol in FY2013. Although, there were no statistically significant differences by ECP user status among adults with diabetes in the percentage with high A1c in FY2013, the difference between FY2011 and FY2013 in the percent with high A1c was lower among ECP users.
Table 3.
Health status and service utilization in fiscal years (FY) 2011 and 2013 by utilization of education, case management, and advanced practice pharmacy (ECP) services in FY2012. Bivariate associations.a
| All adults with diabetes | Adults with diabetes absent CVD | Adults with diabetes and CVD | |||||||
|---|---|---|---|---|---|---|---|---|---|
| ECP non-users | ECP users | ECP non-users | ECP users | ECP non-users | ECP users | ||||
| All adults with diabetes | 16,855 | 11,723 | 12,733 | 8,044 | 4,122 | 3,679 | |||
| Clinical Measures | |||||||||
| High SBP ≥ 140 mmHgb | |||||||||
| FY 2011 | 23.6% | 22.6% | 21.8% | 20.4% | 29.0% | 27.5% | |||
| FY 2013 | 25.9% | 23.8% | ** | 24.4% | 21.5% | *** | 30.7% | 28.7% | |
| Difference | 2.4% | 1.2% | 2.6% | 1.2% | 1.7% | 1.2% | |||
| High hemoglobin A1c ≥ 8%b | |||||||||
| FY 2011 | 42.5% | 44.5% | * | 43.2% | 45.4% | * | 40.4% | 42.5% | |
| FY 2013 | 46.3% | 46.7% | 47.2% | 47.5% | 43.8% | 45.0% | |||
| Difference | 3.9% | 2.2% | * | 4.0% | 2.1% | * | 3.4% | 2.4% | |
| High LDL cholesterol ≥ 100 mg/dLb | |||||||||
| FY 2011 | 44.1% | 38.5% | *** | 46.9% | 42.0% | *** | 36.2% | 31.5% | ** |
| FY 2013 | 38.8% | 34.0% | *** | 41.3% | 37.6% | *** | 31.8% | 26.7% | ** |
| Difference | −5.3% | −4.6% | −5.6% | −4.5% | −4.3% | −4.7% | |||
| Onset of CVD 2013 | 7.3% | 9.1% | *** | ||||||
| Onset of ESRD 2013 | 0.5% | 0.4% | 0.2% | 0.1% | 1.5% | 1.1% | |||
| Hospital service use | |||||||||
| Average number of I/T ED visits | |||||||||
| FY 2011 | 0.79 | 0.78 | 0.69 | 0.62 | ** | 1.10 | 1.12 | ||
| FY 2013 | 0.82 | 0.75 | *** | 0.75 | 0.64 | *** | 1.03 | 0.98 | |
| Difference | 0.03 | −0.03 | ** | 0.07 | 0.02 | * | −0.06 | −0.14 | |
| Percent with 1 or more hospitalizations | |||||||||
| FY 2011 | 8.8% | 10.3% | *** | 5.5% | 5.7% | 19.9% | 21.2% | ||
| FY 2013 | 8.6% | 8.9% | 6.8% | 6.5% | 14.8% | 14.5% | |||
| Difference | −0.2% | −1.4% | ** | 1.3% | 0.8% | −5.2% | −6.7% | ||
| Average number of inpatient days | |||||||||
| FY 2011 | 0.49 | 0.61 | ** | 0.23 | 0.25 | 1.32 | 1.46 | ||
| FY 2013 | 0.62 | 0.63 | 0.41 | 0.36 | 1.28 | 1.27 | |||
| Difference | 0.13 | 0.02 | 0.18 | 0.11 | 0.04 | −0.19 | |||
| Percent with 1 or more I/T potentially preventable hospitalizations | |||||||||
| FY 2011 | 2.1% | 2.6% | ** | 1.1% | 1.3% | 5.0% | 5.5% | ||
| FY 2013 | 2.3% | 2.3% | 1.5% | 1.4% | 4.7% | 4.2% | |||
| Difference | 0.2% | −0.4% | * | 0.4% | 0.1% | −0.4% | −1.3% | ||
FY: fiscal year; CVD: cardiovascular disease; SBP: systolic blood pressure; LDL: low-density lipoprotein; ESRD: end-stage renal disease; I/T: Indian Health Service and Tribal
p < 0.05;
p < 0.01;
p < 0.001
Associations between values for ECP users and non-users during FY2011, FY2013, and the difference between the 2 years.
These clinical measures were assessed for adults in the study population who had both FY2011 and FY2013 data.
Changes in hospital service utilization between FY2011 and FY2013 also differed by ECP user status. Hospitalization and potentially preventable hospitalization results were similar among all adults with diabetes. In FY2011, a higher percentage of ECP users, compared to non-users, had 1 or more hospitalizations (10.3% compared to 8.8%, p<0.001) and 1 or more potentially preventable hospitalizations (2.6% compared to 2.1%, p<0.01). ECP users, compared to non-users, had statistically significant decreases between FY2011 and FY2013 in both measures.
IPTW regression results on the associations of any ECP use and level of ECP use with patient outcomes are summarized in Tables 4 and 5, respectively. Among all adults with diabetes, use of ECP services, compared to no use, was associated with lower odds of having high SBP (OR=0.85, p<0.001) and high LDL cholesterol (OR=0.89, p<0.01) in FY2013. There was no statistical relationship between any use of ECP services and high A1c. Among adults absent CVD, there was no statistically significant association between any use of ECP, compared to no use, and onset of CVD during FY2013. Due to very low ESRD onset rates in FY2013 among all adults with diabetes and adults with diabetes absent CVD, we only evaluated the relationship between ECP use and ESRD onset among adults with diabetes and CVD. ECP users compared to non-users were found to have lower odds of ESRD onset during FY2013 (OR=0.60, p<0.05).
Table 4.
Patient outcomes during fiscal year (FY) 2013 associated with use of education, case management, and advanced practice pharmacy services during FY2012. Propensity score model, based on a logistic regression, using inverse probability weighting.a
| All adults with diabetes | Adults with diabetes absent CVD | Adults with diabetes and CVD | ||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Patient outcome | Sample sizeb | Average Treatment Effect | Odds Ratio (OR) | OR Confidence Interval unless indicatedc | Sample sizeb | Average Treatment Effect | Odds Ratio (OR) | OR Confidence Interval unless indicatedc | Sample sizeb | Average Treatment Effect | Odds Ratio (OR) | OR Confidence Interval unless indicatedc | ||||||
| Health status: Clinical measures | ||||||||||||||||||
| High SBP (≥140mmHg)d | 18,557 | −2.5% | *** | 0.85 | *** | (0.79, 0.93) | 13,659 | −2.8% | *** | 0.83 | *** | (0.75, 0.91) | 4,898 | −1.3% | 0.93 | (0.81, 1.08) | ||
| High HbA1c (≥8%)d | 16,907 | 1.4% | 1.08 | (1.00, 1.16) | 12,323 | 1.3% | 1.07 | (0.98, 1.17) | 4,584 | 1.8% | 1.11 | (0.95, 1.28) | ||||||
| High LDL cholesterol (≥100mg/dL)d | 14,813 | −2.3% | ** | 0.89 | ** | (0.84, 0.98) | 10,721 | −1.8% | 0.92 | (0.84, 1.01) | 4,092 | −3.5% | * | 0.82 | * | (0.71, 0.96) | ||
| Health status: Onset of comorbidities | ||||||||||||||||||
| CVDd.e | 18,869 | −0.8% | 0.89 | (0.78, 1.00) | ||||||||||||||
| ESRDd,e | 7,638 | −0.7% | * | 0.60 | * | (0.39, 0.93) | ||||||||||||
| Hospital service utilization during FY2013 | ||||||||||||||||||
| Emergency visitsf | 27,419 | −0.08 | *** | - | (−0.12, −0.05) | 19,945 | −0.09 | *** | - | (−0.13, −0.05) | 7,474 | −0.10 | * | (−0.19, −0.02) | ||||
| 1 or more hospitalizationsd | 22,456 | −1.4% | *** | 0.80 | *** | (0.71, 0.89) | 16,664 | −1.4% | *** | 0.77 | *** | (0.66, 0.89) | 5,792 | −2.7% | ** | 0.74 | ** | (0.61, 0.91) |
| 1 or more potentially preventable hospitalizationsd | 27,419 | −0.5% | * | 0.79 | * | (0.64, 0.91) | 19,945 | −0.4% | 0.77 | (0.59, 1.00) | 7,474 | −1.3% | ** | 0.71 | ** | (0.55, 0.91) | ||
| Hospital inpatient daysf | 22,456 | −0.13 | ** | - | (−0.21, −0.04) | 16,664 | −0.13 | *** | - | (−0.20, −0.06) | 5,792 | −0.32 | * | (−0.57, −0.07) | ||||
CVD: cardiovascular disease; SBP: systolic blood pressure; IHS: Indian Health Service; ESRD: end stage renal disease
p < 0.05;
p < 0.01;
p < 0.001
All treatment and outcome models are adjusted for baseline (FY 2011) age, gender, health coverage, comorbidities, risk score, county-level education and income, systolic blood pressure ≥140 mmHg, and A1c ≥8%. Models were also adjusted for ECP drive time and ECP facility rate in FY 2012.
Sample sizes for analyses within one condition group varied by outcome assessed based on site exclusions due either to the provision of services (i.e., the site did not provide inpatient services) or data quality.
95% confidence intervals are for odds ratios (ORs) except for 2 outcomes (i.e., emergency visits and hospital inpatient days) for which no ORs were reported. For these outcomes, the 95% confidence intervals are for the average treatment effects.
The regressions were estimated using a logistic regression. The estimated average treatment effect may be interpreted as a change in the probability of being positive for the outcome.
Analyses for the onset of CVD and ESRD in 2013 were limited to persons without diagnostic codes for those conditions in either FY2011 or FY2012.
The regressions were estimated using a negative binomial model. The estimated average treatment effect may be interpreted as an absolute change in the expected count.
Table 5.
Patient outcomes during fiscal year (FY) 2013 associated with level of education, case management, and advanced practice pharmacy (ECP) service use during FY2012. Propensity score model, based on an ordered logistic regression, using inverse probability weighting.a
| All adults with diabetes | Adults with diabetes absent CVD | Adults with diabetes and CVD | |||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Patient outcome | 1–2 ECP visits (vs 0 visits) | 3+ ECP visits (vs 0 visits) | 3+ ECP visits (vs 1–2 visits) | 1–2 ECP visits (vs 0 visits) | 3+ ECP visits (vs 0 visits) | 3+ ECP visits (vs 1–2 visits) | 1–2 ECP visits (vs 0 visits) | 3+ ECP visits (vs 0 visits) | 3+ ECP visits (vs 1–2 visits) | ||||||||||||||||||
| Health status: Clinical measures | |||||||||||||||||||||||||||
| High systolic blood pressure (≥140 mmHg)b | |||||||||||||||||||||||||||
| Average Treatment Effect | −1.8% | * | −4.8% | *** | −3.0% | * | −2.4% | ** | −4.6% | ** | −2.2% | −0.4% | −3.0% | −2.6% | |||||||||||||
| Odds Ratio | 0.89 | * | 0.74 | *** | 0.82 | * | 0.85 | ** | 0.73 | ** | 0.86 | 0.98 | 0.85 | 0.87 | |||||||||||||
| OR Confidence Interval | (0.82, 0.97) | (0.63, 0.86) | (0.70, 0.97) | (0.77, 0.95) | (0.60, 0.89) | (0.69, 1.06) | (0.84, 1.15) | (0.68, 1.07) | (0.67, 1.12) | ||||||||||||||||||
| High A1c (≥8%)b | |||||||||||||||||||||||||||
| Average Treatment Effect | 1.7% | * | 0.4% | −1.3% | 1.8% | −1.5% | −3.3% | 0.9% | 4.5% | * | 3.6% | ||||||||||||||||
| Odds Ratio | 1.09 | * | 1.02 | 0.93 | 1.10 | 0.93 | 0.84 | 1.05 | 1.28 | * | 1.22 | ||||||||||||||||
| OR Confidence Interval | (1.004, 1.186) | (0.89, 1.17) | (0.80, 1.09) | (1.00, 1.21) | (0.78, 1.10) | (0.70, 1.02) | (0.89, 1.24) | (1.02, 1.61) | (0.94, 1.57) | ||||||||||||||||||
| High LDL cholesterol (≥100 mg/dL)b | |||||||||||||||||||||||||||
| Average Treatment Effect | −1.4% | −5.1% | *** | −3.7% | * | −0.7% | −6.3% | *** | −5.6% | ** | −3.7% | * | −2.5% | 1.1% | |||||||||||||
| Odds Ratio | 0.93 | 0.77 | *** | 0.83 | * | 0.97 | 0.73 | *** | 0.75 | ** | 0.82 | * | 0.87 | 1.07 | |||||||||||||
| OR Confidence Interval | (0.86, 0.14) | (0.62, 0.00) | (0.72, 0.04) | (0.87, 1.07) | (0.61, 0.87) | (0.62, 0.91) | (0.68, 0.97) | (0.68, 1.11) | (0.81, 1.40) | ||||||||||||||||||
| Health status: Onset of comorbidities | |||||||||||||||||||||||||||
| CVDb,c | |||||||||||||||||||||||||||
| Average Treatment Effect | −0.6% | −1.5% | * | −1.0% | |||||||||||||||||||||||
| Odds Ratio | 0.92 | 0.79 | * | 0.85 | |||||||||||||||||||||||
| OR Confidence Interval | (0.80, 1.05) | (0.63, 0.99) | (0.67, 1.09) | ||||||||||||||||||||||||
| ESRDb,c | |||||||||||||||||||||||||||
| Average Treatment Effect | −0.8% | * | −0.6% | 0.2% | |||||||||||||||||||||||
| Odds Ratio | 0.55 | * | 0.64 | 1.16 | |||||||||||||||||||||||
| OR Confidence Interval | (0.33, 0.92) | (0.32, 1.29) | (0.55, 2.48) | ||||||||||||||||||||||||
| Hospital service utilization during FY2013 | |||||||||||||||||||||||||||
| Emergency visitsd | |||||||||||||||||||||||||||
| Average Treatment Effect | −0.08 | *** | −0.11 | ** | −0.02 | −0.08 | ** | −0.13 | ** | −0.05 | −0.14 | ** | −0.05 | 0.08 | |||||||||||||
| ATE Confidence Interval | (−0.13, −0.04) | (−0.17, −0.04) | (−0.09, 0.05) | (−0.12, −0.04) | (−0.19, −0.06) | (−0.12, 0.03) | (−0.24, −0.04) | (−0.19, 0.08) | (−0.06, 0.23) | ||||||||||||||||||
| 1 or more hospitalizationsb | |||||||||||||||||||||||||||
| Average Treatment Effect | −1.3% | ** | −1.8% | ** | −0.6% | −1.2% | ** | −1.8% | * | −0.6% | −2.8% | * | −3.1% | * | −0.3% | ||||||||||||
| Odds Ratio | 0.82 | ** | 0.75 | ** | 0.91 | 0.80 | ** | 0.70 | * | 0.88 | 0.75 | * | 0.72 | * | 0.97 | ||||||||||||
| OR Confidence Interval | (0.72, 0.93) | (0.61, 0.91) | (0.73, 1.13) | (0.68, 0.94) | (0.53, 0.93) | (0.65, 1.19) | (0.60, 0.93) | (0.53, 0.98) | (0.69, 1.37) | ||||||||||||||||||
| 1 or more potentially preventable hospitalizationsb | |||||||||||||||||||||||||||
| Average Treatment Effect | −0.3% | −0.9% | ** | −0.6% | * | −0.3% | −0.5% | * | −0.2% | −0.8% | −2.1% | ** | −1.3% | * | |||||||||||||
| Odds Ratio | 0.86 | 0.62 | ** | 0.72 | * | 0.80 | 0.65 | * | 0.81 | 0.81 | 0.54 | ** | 0.667 | * | |||||||||||||
| OR Confidence Interval | (0.70, 1.05) | (0.46, 0.82) | (0.53, 0.98) | (0.60, 1.07) | (0.420, 0.997) | (0.51, 1.29) | (0.61, 1.08) | (0.37, 0.79) | (0.447, 0.996) | ||||||||||||||||||
| Hospital inpatient daysd | |||||||||||||||||||||||||||
| Average Treatment Effect | −0.09 | −0.23 | *** | −0.14 | * | −0.11 | ** | −0.20 | *** | −0.10 | −0.34 | * | −0.41 | * | −0.08 | ||||||||||||
| ATE Confidence Interval | (−0.19, 0.01) | (−0.34, −0.12) | (−0.26, −0.01) | (−0.18, −0.03) | (−0.30, −0.11) | (−0.19, 0.00) | (−0.61, −0.06) | (−0.74, −0.08) | (−0.40, 0.25) | ||||||||||||||||||
CVD: cardiovascular disease; OR: odds ratio; ATE: average treatment effect
p < 0.05;
p < 0.01;
p < 0.001; 95% confidence intervals
Sample sizes for analyses varied by outcome assessed based on site exclusions due either to the provision of services (i.e., the site did not provide emergency services) or data quality.
The regressions were estimated using a logistic regression. The estimated average treatment effect may be interpreted as a change in the probability of being positive for the outcome.
Analyses for the onset of CVD and ESRD in 2013 were limited to persons without diagnostic codes for those conditions in either FY2011 or FY2012.
The regressions were estimated using a negative binomial model. The estimated average treatment effect may be interpreted as an absolute change in the expected count.
Among all adults with diabetes, ECP users compared to non-users had a statistically lower average number of ED visits (−0.08, p<0.01), and lower odds of 1 or more hospitalizations (0.80, p<0.001) and 1 or more potentially preventable hospitalizations (0.79, p<0.05). ECP use was significantly associated with fewer hospital inpatient days (−0.13, p<0.01).
We examined the relationships between ECP use, both any use and level of use, and patient outcomes for the 7 patient outcomes assessed for adults with and without CVD. The relationships between ECP use and patient outcomes was alike (i.e., significant association with improvement in patient outcome, no association) across these two health status groups for 4 of the 7 outcomes examined (i.e., high A1c, ED visits, 1 or more hospitalizations, hospital inpatient days).
The level of ECP use was significantly associated with improved patient outcomes for 5 of the 9 outcomes assessed (e.g., high SBP, high LDL cholesterol, onset of CVD, potentially preventable hospitalizations, hospital inpatient days; Table 5). Among all adults with diabetes, 1–2 ECP visits and 3 or more ECP visits, compared to no visits, were each associated with lower odds of high SBP (OR=0.89 [p<0.05] and OR=0.74 [p<0.001], respectively), and 3 or more ECP visits was associated with lower odds of high SBP than 1–2 visits (OR=0.82, p<0.05). Similarly, for high LDL cholesterol, adults with diabetes who had 3 or more visits, as compared to no visits, and patients with 3 or more visits, as compared to 1–2 visits, had lower odds of high LDL cholesterol (OR=0.77 [p<0.001] and OR=0.83 [p<0.05], respectively).
ECP use, compared to no use, was not associated with lower onset of CVD or lower odds of 1 or more potentially preventable hospitalizations among adults absent CVD. However, among these adults, those who had 3 or more ECP visits, compared to those with no visits, had lower odds of CVD onset (OR=0.79, p<0.05) and lower odds of 1 or more potentially preventable hospitalizations (OR=0.65, p<0.05). Among all adults with diabetes, patients who had 3 or more visits, as compared to no visits, and patients who had 3 or more visits, as compared to 1–2 visits, had fewer hospital inpatient days (−0.23, [p<0.001] and −0.14 [p<0.05], respectively).
Discussion
This is the first large scale study, to our knowledge, to evaluate patient outcomes associated with ECP utilization among AI/ANs with diabetes. We found ECP use was associated with improvements in blood pressure and cholesterol control, lower odds of CVD and ESRD onset, and reductions in hospital emergency department and inpatient utilization. While we did not observe an association between ECP use and glycemic control in FY2013, there were a number of factors that could have influenced this finding, including a larger percentage of ECP users having high A1c at baseline (FY2011). Due to the importance of controlling blood sugar levels, future research on provider referrals to ECP and how blood sugar control is addressed during ECP and other outpatient visits is warranted.
Despite differences in age and morbidity burden between adults with diabetes absent CVD and with CVD, we found similar relationships between ECP use and patient outcomes across these 2 health status groups for 4 of the 7 outcomes examined. In addition, the findings suggest that higher levels of ECP use (i.e., 3 or more ECP visits) were associated with improved outcomes for 5 of the 9 outcomes assessed.
Our study has limitations that merit consideration. First, due to the observational design, residual confounding related to patient self-selection to ECP use could account for some of the observed associations. Propensity score models, such as IPTW regression, control for observable confounders (e.g., health status, drive time to services), but there may have been important unobserved provider and patient characteristics associated with ECP use and patient outcomes that could have biased our results. For example, coordination between ECP and primary care providers may vary across project sites, with higher coordination levels likely benefiting patients. Patient motivation to maintain or improve one’s health status may not only influence ECP use but also patient outcomes. Thus, lacking measures of site coordination, patient motivation, and other confounding factors, we may have overestimated the influence of ECP use on patient outcomes. Future quasi-experiment or randomized clinical trials designed to test the effects of ECP use among AI/ANs are needed to validate our findings.
Many previous ECP effectiveness studies evaluated interventions designed to provide more than 3 visits during a time period, typically a 12-month period.14–21 For example, SDPI Healthy Heart demonstration project participants had, on average, 7 case management visits during the first program year, and those with more visits experienced significantly greater improvements in some outcomes.23 The average number of ECP visits among ECP users in this study was 2.8. We assessed ECP use during the 12 months of FY2012 and did not account for ECP use prior to or after FY2012. We may assess ECP use during individually defined time periods in future studies. Although we employed an algorithm to identify ECP visits that allowed for project site specific adjustments, we may not have identified all provided ECP visits. Finally, we assessed outcomes during a 12-month follow-up period, and the results concerning onset of CVD and ESRD were borderline (i.e., p<0.05). Longer time periods could be used to understand the extent to which ECP use may mitigate the onset of these conditions.
Other study limitations pertain to the nature of the IHS Data Project data. We reported the prevalence of conditions based on diagnoses included in medical service utilization records. While this method allowed us to include a large number of AI/ANs in the analysis, we did not have the detail and accuracy of medical records. These data include information for services provided by I/T programs or were paid by the PRC program. We did not have data on other services used by the study population, and financial and geographic access to other services (e.g., non-I/T specialty services) varied across project sites. This limitation may have biased downward morbidity measures, a bias which likely varied across the sites. Sites also varied by the types of services provided (e.g., specialty outpatient and inpatient services), PRC service utilization, funding and completeness of data. While software issues contributed to missing data at some sites, there may have been a relationship between missing data and quality of care. It is difficult to predict the influence of these limitations on study findings. An ECP cost-effectiveness analysis was beyond the scope of this study. However, our subsequent work will assess ECP costs, savings, and cost-effectiveness. Lastly, our study population represents a large proportion of AI/ANs eligible for I/T health services. Nevertheless, findings may not reflect the health status of AI/AN peoples who live elsewhere or who do not obtain health services from I/T providers.11,33
Despite these limitations, we were able to analyze existing IHS/Tribal electronic health data to evaluate ECP utilization for a large and geographically representative sample of adults with diabetes. Findings on positive patient outcomes associated with ECP utilization by adults with diabetes may inform IHS/Tribal policies, funding, and enhancements to ECP services and improve patient knowledge of ECP services, and ultimately contribute to reducing disparities between AI/ANs and other populations in diabetes-related morbidity and mortality.
Supplementary Material
Acknowledgements:
The data used in this study are from the Indian Health Service (IHS) Health Care Delivery Data Project. The data include information from many American Indian and Alaska Native communities. Data Project analyses are conducted with the guidance and advice of IHS and Tribal health program colleagues, as well as members of the project’s Steering, Project Site, and Patient Committees. Members of Tribal and IHS institutional review boards, Tribal Councils, and Tribal Authorities educate us about the health concerns they have for their Tribal members and how they hope this project will inform their work. This project relies on their support and approval. The authors would like to express their gratitude to Sara Mumby for her editorial assistance.
Disclosure of funding:
The research reported in this publication was supported the Patient-Centered Outcomes Research Institute (AD-1304-6451), National Institutes of Health (NIH) National Institute of Diabetes and Digestive and Kidney Diseases (R18DK114757 and P30DK092923), and the NIH National Institute on Aging (R01AG061189 and P30AG15292). Funding for the development of the data infrastructure, utilized in the reported analyses, was supported by the Agency for Healthcare Research and Quality (290-2006-00020-I, TO #11, J.M. O’Connell). The content of this report is solely the responsibility of the authors and does not necessarily represent the official views of these organizations.
Footnotes
Disclosure of conflicts of interest: All authors report that no potential conflicts of interest exist.
Contributor Information
Joan O’Connell, Centers for American Indian and Alaska Native Health, Colorado School of Public Health, University of Colorado Anschutz Medical Campus, 13055 East 17th Avenue | Aurora, CO 80045.
Margaret Reid, Department of Health Systems Management and Policy, Colorado School of Public Health, University of Colorado Anschutz Medical Campus, 13055 East 17th Avenue | Aurora, CO 80045.
Jennifer Rockell, formerly at UCD CAIANH, Telligen, 7730 E Belleview Ave | Suite 300 | Greenwood Village, CO 80111.
Kathleen Harty, Jefferson Center for Mental Health, 4851 Independence St. | Wheat Ridge, CO, 80033.
Marcelo Perraillon, Department of Health Systems Management and Policy, Colorado School of Public Health, University of Colorado Anschutz Medical Campus, 13001 East 17th Place | Aurora, CO 80045.
Spero Manson, Centers for American Indian and Alaska Native Health, Colorado School of Public Health, University of Colorado Anschutz Medical Campus, 13055 East 17th Avenue | Aurora, CO 80045.
References
- 1.Centers for Disease Control and Prevention. National Diabetes Statistics Report, 2020. Atlanta, GA: Centers for Disease Control and Prevention, U.S. Department of Health and Human Services;2020. [Google Scholar]
- 2.Espey DK, Jim MA, Cobb N, et al. Leading causes of death and all-cause mortality in American Indians and Alaska Natives. Am J Public Health. 2014;104(S3):S303–S311. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Barnes PM, Adams PF, Powell-Griner E. Health Characteristics of the American Indian and Alaska Native Adult Population: United States, 1999–2003. Rockville, MD: U.S. Department of Health and Human Services, National Center for Health Statistics;2005. [PubMed] [Google Scholar]
- 4.O’Connell J, Yi R, Wilson C, et al. Racial disparities in health status: A comparison of the morbidity among American Indian and U.S. adults with diabetes. Diabetes Care. 2010;33:1–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.U.S. Department of Health and Human Services, Indian Health Service. Fact sheet: Special Diabetes Program for Indians (SDPI), Changing the course of diabetes. 2017. Available at: https://www.ihs.gov/sdpi/includes/themes/responsive2017/display_objects/documents/factsheets/SDPI_FactSheet_July2017.pdf. Accessed 04/29/2019.
- 6.Cho P, Geiss LS, Burrows NR, et al. Diabetes-related mortality among American Indians and Alaska Natives, 1990–2009. Am J Public Health. 2014;104(S3):S496–S503. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Veazie M, Ayala C, Schieb L, et al. Trends and disparities in heart disease mortality among American Indians/Alaska Natives, 1990–2009. Am J Public Health. 2014;104(S3):S359–S367. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Schieb LJ, Ayala C, Valderrama AL, et al. Trends and disparities in stroke mortality by region for American Indians and Alaska Natives. Am J Public Health. 2014;104(S3):S368–376. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.U. S. Department of Health and Human Services IHS. IHS Profile. Based on 2015–2019 data -- Numbers are approximate. U.S. Department of Health and Human Services. 2020. Available at: https://www.ihs.gov/newsroom/factsheets/ihsprofile/. Accessed 06/16/2020. [Google Scholar]
- 10.Indian Health Service, Division of Diabetes Treatment and Prevention. Diabetes Standards of Care & Clinical Practice Resources. Indian Health Service. 2020. Available at: https://www.ihs.gov/diabetes/clinician-resources/soc/. Accessed 06/16/2020. [Google Scholar]
- 11.Indian Health Service, Special Diabetes Program for Indians. 2014 Report to Congress. Changing the Course of Diabetes: Turning Hope into Reality. 2014. [PubMed] [Google Scholar]
- 12.Agency for Healthcare Research and Quality. Hospital admissions for uncontrolled diabetes improving among American Indians and Alaska Natives. Rockville, MD: U.S. Department of Health and Human Services;2018. No. 18(19)-0033–7-EF. [Google Scholar]
- 13.U.S. Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation. The Special Diabetes Program for Indians: Estimates of Medicare savings. ASPE Issue Brief. 2019:1–29. https://aspe.hhs.gov/pdf-report/special-diabetes-program-indians-estimates-medicare-savings. Accessed 7/24/2019. [Google Scholar]
- 14.Yu J, Shah BM, Ip EJ, et al. A Markov model of the cost-effectiveness of pharmacist care for diabetes in prevention of cardiovascular diseases: evidence from Kaiser Permanente Northern California. J Manag Care Pharm. 2013;19(2):102–114. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Look ARG, Wing RR. Long-term effects of a lifestyle intervention on weight and cardiovascular risk factors in individuals with type 2 diabetes mellitus: four-year results of the Look AHEAD trial. Arch Intern Med. 2010;170(17):1566–1575. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Espeland MA, Glick HA, Bertoni A, et al. Impact of an intensive lifestyle intervention on use and cost of medical services among overweight and obese adults with type 2 diabetes: the action for health in diabetes. Diabetes Care. 2014;37(9):2548–2556. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Heisler M, Hofer TP, Schmittdiel JA, et al. Improving blood pressure control through a clinical pharmacist outreach program in patients with diabetes mellitus in 2 high-performing health systems: the adherence and intensification of medications cluster randomized, controlled pragmatic trial. Circulation. 2012;125(23):2863–2872. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Santschi V, Chiolero A, Paradis G, et al. Pharmacist interventions to improve cardiovascular disease risk factors in diabetes: a systematic review and meta-analysis of randomized controlled trials. Diabetes Care. 2012;35(12):2706–2717. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.CDC Diabetes Cost-effectiveness Group. Cost-effectiveness of intensive glycemic control, intensified hypertension control, and serum cholesterol level reduction for type 2 diabetes. JAMA. 2002;287(19):2542–2551. [DOI] [PubMed] [Google Scholar]
- 20.Gregg EW, Chen H, Wagenknecht LE, et al. Association of an intensive lifestyle intervention with remission of type 2 diabetes. JAMA. 2012;308(23):2489–2496. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Nuckols TK, Keeler E, Anderson LJ, et al. Economic Evaluation of Quality Improvement Interventions Designed to Improve Glycemic Control in Diabetes: A Systematic Review and Weighted Regression Analysis. Diabetes Care. 2018;41(5):985–993. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Roubideaux Y, Noonan C, Goldberg JH, et al. Relation between the level of American Indian and Alaska Native diabetes education program services and quality-of-care indicators. Am J Public Health. 2008;98(11):2079–2084. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Moore K, Jiang LH, Manson SM, et al. Case Management to Reduce Cardiovascular Disease Risk in American Indians and Alaska Natives With Diabetes: Results From the Special Diabetes Program for Indians Healthy Heart Demonstration Project. Am J Public Health. 2014;104(11):E158–E164. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.O’Connell J, Guh S, Ouellet J, et al. ARRA ACTION: Comparative Effectiveness of Health Care Delivery Systems for American Indians and Alaska Natives Using Enhanced Data Infrastructure: Final Report. Rockville, MD: Agency for Healthcare Research and Quality;2014. [Google Scholar]
- 25.Sequist TD, Cullen T, Acton KJ. Indian health service innovations have helped reduce health disparities affecting American Indian and Alaska Native people. Health Aff (Millwood). 2011;30(10):1965–1973. [DOI] [PubMed] [Google Scholar]
- 26.Centers for Disease Control and Prevention, Office for State T, Local, and Territorial Support. Public Health Practice Stories from the Field. Pharmacists help improve health of Yakama Indians living with diabetes. 4/2/2015 2015. [Google Scholar]
- 27.U.S. Department of Health and Human Services, Center for Medicare and Medicaid Services. National Health Expenditure Accounts 2018. Available at: https://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/NationalHealthExpendData/NationalHealthAccountsHistorical.html. Updated2018 Accessed 05/07/2020. [Google Scholar]
- 28.United States Government Accountability Office. Report to the Committee on Indian Affairs, U.S. Senate. Indian Health Service. Actions needed to improve oversight of patient wait times. March 2016. [Google Scholar]
- 29.Zuckerman S, Haley J, Roubideaux Y, et al. Health service access, use, and insurance coverage among American Indians/Alaska Natives and Whites: What role does the Indian Health Service play? Am J Public Health. 2004;94(1):53–59. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Lillie-Blanton M, Roubideaux Y. Understanding and addressing the health care needs of American Indians and Alaska Natives. Am J Public Health. 2005;95(5):759–761. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Giberson S, Yoder S, Lee MP. Improving patient and health system outcomes through Advanced Pharmacy Practice. A report to the U.S. Surgeon General. Office of the Chief Pharmacist, U.S. Public Health Service; Dec 2011. [Google Scholar]
- 32.U.S. Government Accountability Office. Indian Health Service: Agency faces ongoing challenges filling provider vacancies. 2018. GAO-18–580.
- 33.U.S. Census Bureau. American Indian and Alaska Native Heritage Month: November 2017. Profile America Facts for Features. 2017. https://www.census.gov/newsroom/facts-for-features/2017/aian-month.html. Accessed 12/9/2019. [Google Scholar]
- 34.Indian Health Service. User Population Report (HQ Summary), Three Year; Age Groups and Sex by Areas and Service Units of Residence, Fiscal Year 2010. Rockville, MD: U.S. Department of Health and Human Services; 1/23/2013 2013. [Google Scholar]
- 35.U.S. Department of Health and Human Services, Indian Health Service, Division of Diabetes Treatment and Prevention. AUDIT 2015 Instructions, IHS Diabetes Care and Outcomes Audit for Quality Assessment and Improvement. Rockville, MD: U.S. Department of Health and Human Services;2015. [Google Scholar]
- 36.Nichols GA, Schroeder EB, Karter AJ, et al. Trends in diabetes incidence among 7 million insured adults, 2006–2011, The SUPREME-DM Project. Am J Epidemiol. 2014;181(1):32–39. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.U.S. Renal Data System. 2017 USRDS Annual Data Report: Epidemiology of kidney disease in the United States. Bethesda, MD: National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases;2017. [Google Scholar]
- 38.Verisk Health, Inc. Sightlines™ DxCG Risk Solutions. Version 4.0.1. In:2011.
- 39.American Diabetes Association. Standards of medical care in diabetes—2015. Diabetes Care. 2015;38(Supplement 1):S1–S89. [Google Scholar]
- 40.Bott AM, Collins J, Daniels-Costa S, et al. Clinical pharmacists improve patient outcomes and expand access to care. Federal Practitioner. 2019;36(10):471–475. [PMC free article] [PubMed] [Google Scholar]
- 41.Lamer C, Maynard L, Mason D. The future of clinical pharmacy services in the IHS. The IHS Primary Care Provider. 2005;30(1):1–5. [Google Scholar]
- 42.National Space and Aeronautics Administration. My Nasa Data, NASA Latitude/Longitude Finder. Available at: mynasadata.larc.nasa.gov/latitudelongitude-finder/. Accessed 8/19/15.
- 43.U.S. Census Bureau, U.S. Department of Commerce. American Community Survey. 2010–2014. Available at: https://www.census.gov/programs-surveys/acs/.Accessed
- 44.SAS software [computer program]. Version 9.4. Cary, NC, USA. 2013. [Google Scholar]
- 45.Stata Statistical Software: Release 14. [computer program]. College Station, TX: StataCorp LP; 2015. [Google Scholar]
- 46.Austin PC, Stuart EA. Moving towards best practice when using inverse probability of treatment weighting (IPTW) using the propensity score to estimate causal treatment effects in observational studies. Statistics in Medicine. 2015;34:3661–3679. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Rosenbaum PR, Rubin DB. The Central Role of the Propensity Score in Observational Studies for Causal Effects. Biometrika. 1983;70(1):41–55. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.