Trends in Health Disparities of Rural Latinos Pre- and Post-Accountable Care Organization Implementation

Abstract

Purpose:

We present findings from a longitudinal investigation, the purpose of which was to compare health disparities of rural Latino older adult patients diagnosed with diabetes to their non-Latino White counterparts.

Methodology/Approach:

A pre-post design was implemented treating Medicare Accountable Care Organization (ACO) participation by Rural Health Clinics (RHCs) as an intervention, and using diabetes-related hospitalizations to measure disparities. Data for a nationwide panel of 2,683 RHCs were analyzed for a study period of eight years: 2008 – 2015. In addition, data were analyzed for a subset of 116 RHCs located in Florida, Texas, and California that participated in a Medicare ACO in one or more years of the study period.

Findings:

Two broad findings resulted from this investigation. First, for both the nationwide panel of RHCs and the three-state sample of “ACO RHCs,” there was a decrease in the mean disparities in diabetes-related hospitalization rates over the eight-year study period. Second, in comparing a three-year time period after Medicare ACO implementation in 2012 to a four-year period before the implementation, a statistically significant difference in mean disparities was found for the nationwide panel.

Research limitations/implications:

There are a number of factors that may contribute to the decrease in diabetes-related hospitalization rates for Latinos in more recent years. Future research will identify specific contributors to reducing diabetes-related hospitalization disparities between Latinos and the general population, including the possible influence of ACO participation by RHCs.

Originality/Value of Paper:

This paper presents original research conducted using data related to rural Latino older adults. The data represent multiple states and an eight-year time period. The U.S. Latino population is growing at a rapid pace. As a group, they are at a high risk for developing diabetes, the complications of which are serious and costly to the patient and the U.S. healthcare system. With the continued growth of the Latino population, it is critical that their health disparities be monitored, and that factors that contribute to their health and well-being be identified and promoted.

Introduction

The Covid-19 pandemic has heightened the awareness that certain sectors of the U.S. population are particularly vulnerable to poor health outcomes. Among the groups that have experienced some of the highest proportions of Covid-19-related cases and deaths is Hispanics/Latinos. As of January 2021, Hispanics/Latinos comprised 20.9% of Covid-19 cases and 13.2% of deaths in all age groups (Centers for Disease Control and Prevention [CDC], 2021.)

Hispanics are the largest rural minority in the U.S. (Probst & Ajmal, 2019), and the fastest growing population in rural America (Lichter, 2012.) By 2060, there will be an estimated 111 million Latinos (rural and urban combined) living in the U.S. (U.S. Census Bureau, 2017.) The term “Hispanic” is used to describe U.S. residents who are of Spanish-speaking background and whose origins are, or descent is from, Mexico, Puerto Rico, Cuba, Central America, South America, or other Spanish-speaking country (Johnson, 2012; McElmurry, McCreary, Park, Ramos, Martinez, Parikh, …& Fogelfeld, 2009.) Recently, the term “Latino” has come into use to describe this group, although the terms “Hispanics” and “Latinos” are often used interchangeably. For purposes of this study, we use the term “Latinos” to describe this population.

The combination of Latino ethnicity and rural residence has far-reaching implications for the provision and costs of healthcare services in the U.S. Latinos are at increased risk for developing diabetes (Koopman, Mainous, & Geesey, 2006), and its complications. The Latino population has higher percentages of overweight or obese persons and lower percentages of insured persons as compared to Blacks/African Americans and Whites (Health, United States, 2017.) These and other factors contribute to the projection that the prevalence of diabetes among Latinos will increase by 149% by 2050 (Engelgua, Geiss, Saaddine, Boyle, Benjamin, Gregg, …& Narayan, 2004.) Latinos living in U.S. rural areas are even more vulnerable to adverse health outcomes. Compared to residents of U.S. urban areas, rural residents are older, have some of the highest rates of adverse health conditions, and experience higher rates of chronic disease such as hypertension (Bolin, Bellamy, Ferdinand, Kash, & Helduser, 2015).

As the U.S. healthcare delivery system shifts from an emphasis on volume to one of value, several new models for healthcare delivery have been developed in an effort to improve health care and health outcomes. One of these models is the Accountable Care Organization (ACO.) While ACOs may be sponsored by either public and private entities, all are designed to incentivize high quality care and lower costs. Medicare ACOs are the focus of this investigation. This type of ACO is defined as “…groups of doctors, hospitals, and other health care providers, who come together voluntarily to give coordinated high-quality care to their Medicare patients (Centers for Medicare and Medicaid Services [CMS], 2020). Of the Medicare ACOs, the most widely adopted model is the Shared Savings Program (SSP) ACO.

While most ACOs are not specifically incentivized to reduce health disparities, we focus on the relationship between ACO participation and health disparities as a desirable outcome of providing good quality care. To assess health disparities of rural Latinos, we examined the trends in hospitalization rates of rural Latino patients diagnosed with diabetes. This investigation was part of a larger study that examined the impact of ACO participation by primary care organizations (as well as other factors) on health disparities of rural Latinos. The overall goal of this study is to contribute to improving service delivery to rural Latino older adult patients.

A health disparity may be defined as “…a health difference that adversely affects disadvantaged populations, based on one or more of the specified health outcomes: higher incidence and/or prevalence of disease and/or disorders…” (2018 Budget, National Institute on Minority Health and Health Disparities, p. 9.) This investigation analyzed the trends in disparities in diabetes-related hospitalizations of rural Latino older adult patients diagnosed with diabetes as compared to their non-Latino White counterparts over an 8-year time period: 2008 – 2015. The time period for the study was selected to achieve another goal of the larger study: to assess the contribution of Rural Health Clinic (RHC) participation in a Medicare Accountable Care Organization (ACO) on reducing disparities in diabetes-related hospitalization rates of rural Latino older adult patients. The eight-year time period includes four years before and three years after the introduction of the Medicare Shared Savings Program (MSSP) ACOs in 2012.

For this investigation we addressed the following research questions:

1. Are there disparities in in diabetes-related hospitalization rates of older rural Latinos as compared to their White counterparts?

2. Was there a reduction in the disparities of diabetes-related hospitalization rates after the introduction of the Medicare Shared Savings Program ACO model?

Related Research

Safety net health facilities in rural and underserved areas are fundamental in providing services to prevent and manage health conditions that might otherwise result in avoidable hospitalizations. Rural Health Clinics (RHCs) and Community Health Centers (CHCs) are categories of safety net facilities that may help to limit Ambulatory Care Sensitive Condition (ACSC) hospitalizations in a given county, particularly for older adults (Probst, Laditka, & Laditka, 2009.)

This investigation analyzed data on older patients served by Rural Health Clinics (RHCs.) RHCs are supported through the Rural Health Clinic Service Act of 1977 to address the inadequate supply of physicians providing care to Medicare beneficiaries in the rural U.S. Today’s approximately 4,500 RHCs provide primary care and preventive services to residents throughout the rural U.S. (Centers for Medicare and Medicaid Services [CMS], 2019.)

Increasingly, RHCs, CHCs, and other primary care organizations throughout the nation are joining ACOs. To date, of the few studies that examine the relationship between ACOs and performance, some report on quality performance using process measures such as preventive care and care coordination (e.g., Lewis, Fraze, Fisher, & Colla, 2018.)

Studies using health outcome measures of Medicare beneficiaries or patients participating in ACOs are limited in number. While some studies examine the impact of ACOs on avoidable hospitalizations, few of these describe the impact of ACOs on health disparities. Moreover, early in the history of ACOs, some researchers speculated that ACOs would not address health disparities of racial and ethnic minorities and other vulnerable groups if providers (Pollack & Armstrong, 2011). Indeed, this viewpoint was supported by some early studies. For example, Anderson and colleagues (2014) examined the quality outcomes of Medicare ACO beneficiaries finding no impact on reducing disparities in preventable hospitalizations. A more recent study analyzed hospitalization data from one healthcare system that operates as an ACO (Yaqoob, Wang, Sweeney, Wells, Rego, & Jaber, 2018). Among their findings were that over the five-year period of 2012 – 2016, the annual hospitalization rate for uncontrolled diabetes did not change for non-Whites. However, the rate increased for older adults, Whites, and those with Medicare coverage.

Methods

Design

This investigation was the first of two phases of a multi-year project, the purpose of which is to: 1) compare health disparities and patient outcomes of rural Latino older adult patients diagnosed with diabetes to their non-Latino White counterparts, and 2) ascertain the impact of ACO participation by rural primary care providers (along with other factors) on rural Latino older adult patients. An extension of the Donabedian (1988) structure-process-outcome framework guided the project. Variable construction and analyses were conducted using SPSS (IBM, 2013), Stata (StataCorp., 2015), SAS (SAS Institute Inc., 2014), and R (The R Project for Statistical Computing, 2015.) The study was approved by the University of Central Florida’s Institutional Review Board.

Population.

The unit of analysis for this investigation was the RHC (or “clinic.”) During the eight-year study period of 2008 – 2015, the number of RHCs nationwide grew from 3,756 to 4,116 (CMS, 2015.) They are located in all states and HHS Regions (Department of Health and Human Services Regions.)

Samples.

Using the Provider of Services data files (Centers for Medicare & Medicaid Services, 2015), we created a panel of 2,683 clinics continuously certified as RHCs for the period 2008 – 2015. The panel contained data on RHCs and their patients. During the study period, some of the RHCs were Medicare ACO participants; others were not. From this panel, we created a subset of 516 RHCs located in three states: California, Florida and Texas. These study states were selected based on several criteria. Each state represents a different Region of the 10 HHS Regions throughout the country. Each had higher numbers of RHCs. In addition, each of the three states had higher numbers of Hispanic/Latinos in their rural areas as compared to many other states. Finally, from the three-state subset of 516 RHCs, we created another subset of 116 RHCs that were Medicare ACO participants during one or more of the study years.

Data Sources and Variables

In this section, we describe the data sources and construction of the variables used in this investigation. Initially, we expanded our existing dataset on RHC characteristics. The resulting dataset contains data for 536 variables which are grouped into the following categories: “context” (sociodemographic and other variables that describe the environment that affects the RHC patients); “structure” (human and material resources that describe the attributes of the healthcare setting); “process” (preventive and management services that describe the primary care services provided); and “outcomes” (the result or effect of the healthcare provided.) The variables were constructed using data from a variety of sources. Context variables were constructed using data from the Area Health Resource Files (AHRF, U.S. Department of Health and Human Services, 2008 – 2015); structure variables were constructed with data from the Medicare Cost Reports (CMS, 2008; CMS 2008a; CMS, 2009 – 2015; CMS, 2009 – 2015a); and the process and outcome variables used data from the CMS Chronic Condition Data Warehouse or CCW (CMS, 2019a.) All data were aggregated to the RHC level.

Study patients were associated with a particular RHC according to where he/she received a plurality of his/her RHC services. Although there is no apparent empirical evidence about the best approach for assigning patients to ACOs, using the “plurality of services” approach has been used by researchers of primary care (e.g., Pope et al., 2014). We controlled for access to other Medicare/Medicaid primary care providers by the inclusion of the variable “physician-population ratio.” This variable was calculated as the ratio of primary care physicians to the population in the county of location of each study RHC.

Variables measuring services provided to patients and patient outcomes were risk-adjusted. RHCs are quite varied in terms of patients they serve and many other factors. We accounted for differences in patient mix using a risk-adjustment process for the outcome variables. For example, hospital admission rates were risk-adjusted using logistic regression analysis where variables such as patient age, gender, comorbidity, and race are included in the model. The rate was computed by dividing an expected number of admissions by the actual number for each RHC for each year.

“Structure” Variable of Interest: ACO.

The structure variables represent organizational structure and other mediating factors, and are calculated on the RHC level. These variables account for aspects of RHC operations under management control. Using dummy variables, we categorized each RHC in the panel as a participant or non-participant in either one of two types of Medicare ACOs: Pioneer ACOs or Medicare Shared Savings Program (MSSP) ACOs (where for each ACO type, 1 = ACO participant and 0 = non-participant.) The data for these variables were combined into one variable, “ACO,” for each study year.

“Outcome” Variables of Interest: Health Disparities.

We used the following definition of “disparity” for this set of variables: “the quantity that separates a group from a specified reference point on a particular measure of health that is expressed in terms of a rate, percentage, mean, or some other quantitative measure (Keppel et al., 2005, p. 3). In constructing the disparities variables, we first constructed a set of patient-related Outcome variables called “ACSC - Diabetes.” This set of variables describes the Ambulatory Care Sensitive Condition (ACSC) Rate for diabetes-related admissions. The variables were measured as the number of discharges among the RHC’s beneficiaries of age 65 and older with diagnosis of Type 2 diabetes divided by the number of beneficiaries with outpatient diagnosis of the condition. The measure was risk-adjusted and accounted for geographic variation across counties and states.

We then constructed both “absolute” and “relative” measures of disparities as recommended by current literature (Duran, Asada, Millum, & Gezmu, 2019; Keppel et al., 2005). Whereas the “absolute” measure of disparity is the arithmetic difference between a group rate and a specified reference point, the “relative” measure is the difference between rates in terms of the reference point. Thus, we calculated the absolute disparity as the difference in diabetes-related hospitalization rates of older Latino beneficiaries served by RHCs as compared to their White counterparts. The relative disparity was calculated as the deviation of the annual rate of Latinos from the reference group’s (Whites) average rate divided by the reference group’s average rate.

Analysis

The first step in the analysis for this investigation was to describe the diabetes-related disparities data for the years 2008 – 2015. Both measures of disparities – absolute and relative – were described using summary statistics such as means, medians, and standard deviations. We then created tables and graphs to report these statistics for each study year. For example, we created tables and corresponding graphs that show the absolute disparities between risk-adjusted, diabetes-related hospitalization rates for rural Latino patients served by “ACO RHCs” (RHCs that participated in either a Pioneer or MSSP ACO), and the risk-adjusted, diabetes-related hospitalization rates for rural non-Latino White patients served by “ACO RHCs” for each study year.

The next step in the analysis was to conduct tests of the disparities in hospitalization rates using statistical tests. The purpose of these tests was to determine: 1) whether the disparities were zero or positive in the population in either the Pre-ACO formation years (2008 – 2011) or the Post-ACO formation years (2013 – 2015); and 2) whether the population disparities were the same or different in the Pre-ACO years and the Post-ACO years.

To determine whether a t-test was appropriate in comparing the mean disparities, we first checked for normality of the Pre- and Post-ACO variables using Q-Q plots and Shapiro-Wilks tests. Neither variable was found to be normally distributed, indicating that a t-test was inappropriate. The next step was to transform the Pre- and Post-ACO variables in multiple ways (square root, natural logarithm, and reciprocal), then check the transformed variables for normality using Q-Q plots and Shapiro-Wilks tests. This step found that the transformed variables were not normally distributed, so using a t-test was deemed inappropriate.

The final step in comparing disparities in hospitalization rates of the Pre- to the Post-ACO years was to employ a binomial method for obtaining confidence intervals that makes no assumptions about the underlying distribution of the variables. In addition, we estimated a regression that also makes no assumptions about the specific underlying distribution of the variables.

Results

From the initial panel of 2,683 clinics continuously certified as RHCs during the study period, we identified 516 RHCs in the three study states (California, Florida, and Texas.) From these, we created a panel of 429 “ACO RHCs” (i.e., RHCs that participated in a Medicare ACO during the study years. Of the “ACO RHCs,” 116 had data that enabled us to construct disparity measures. That is, for many RHCs, disparities variables could not be calculated if the clinic either served no older Latino patients who were hospitalized for diabetes-related conditions, or served no older White patients with such hospitalizations.

Absolute Disparity Results

Table 1 presents statistics for the initial panel of 2,683 clinics located throughout the U.S. These statistics describe the absolute diabetes-related disparities between Latinos and Whites for the panel for each of the study years. Over the study period, the mean disparity declined from 0.009 in 2008 to −0.014 in 2015.

Table 1.

Descriptive Statistics for Absolute Disparities, Nationwide Panel: 2008 – 2015 (n = 2,683)

	2008	2009	2010	2011	2012	2013	2014	2015
Mean	.009262	−.006064	−.005282	−.002658	−.003829	−.008933	−.014141	−.013797
Median	.000221	−.007352	−.007731	−.004383	−.006856	−.009873	−.013941	−.015709
Standard Deviation	.050270	.019578	.018154	.016788	.021556	.015034	.014595	.018569

Note. Values were calculated for the absolute difference between diabetes-related hospitalizations for older rural Latino and White Medicare beneficiaries in a panel of Rural Health Clinics located in all states, 2008 – 2015.

Figure 1 illustrates the trend in absolute diabetes-related disparities for Latinos in the nationwide panel as compared to the reference group – Whites. At the beginning of the study period (2008), Latino patients had a higher number of diabetes-related hospitalizations than their White counterparts. Our results indicate, however, that from 2009 to 2015 Latino patients had slightly lower rates of these hospitalizations as compared to Whites.

Figure 1. Trend in Mean Absolute Disparities, Nationwide Panel: 2008 – 2015 (n = 2,683).

Note. Trend line illustrates the mean absolute difference between diabetes-related hospitalizations for older rural Latino and White Medicare beneficiaries in a panel of Rural Health Clinics located in all states, 2008 – 2015.

Table 2 presents statistics for the 116 “ACO RHCs” in California, Florida, and Texas for which there were data to calculate absolute disparities. This table describes the absolute diabetes-related disparities between Latinos and Whites for the “ACO RHCs” for each of the study years. Over the study period, the mean disparity declined from 0.008 in 2008 to −0.013 in 2015.

Table 2.

Descriptive Statistics for Absolute Disparities, “ACO RHCs,” CA, FL, TX: 2008 – 2015 (n = 116)

	2008	2009	2010	2011	2012	2013	2014	2015
Mean	.008054	−.005211	−.003043	−.003369	−.004509	−.010645	−.014065	−.012979
Median	.005940	−.005327	−.002960	−.004383	−.005488	−.010198	−.013881	−.013570
Standard Deviation	.048783	.014489	.016092	.013170	.014160	.012323	.012302	.014755

Note. Values were calculated for the absolute difference between diabetes-related hospitalizations for older rural Latino and White Medicare beneficiaries in a panel of Rural Health Clinics participating in Medicare ACOs and located in California, Florida, and Texas, 2008 – 2015.

Figure 2 illustrates the trend in absolute diabetes-related disparities for Latinos served by the “ACO RHCs” in the three-state sample as compared to the reference group: Whites. At the beginning of the study period (2008), Latino patients had a higher number of diabetes-related hospitalizations than their White counterparts. Our results indicate that from 2009 to 2015 Latino patients had slightly lower rates of these hospitalizations as compared to Whites.

Figure 2. Trend in Absolute Disparities, “ACO RHCs,” CA, FL, TX: 2008 – 2015 (n = 116).

Note. Trend line illustrates the mean absolute difference between diabetes-related hospitalizations for older rural Latino and White Medicare beneficiaries in a panel of Rural Health Clinics participating in Medicare ACOs and located in California, Florida, and Texas, 2008 – 2015.

Relative Disparity Results

Table 3 presents statistics describing the 116 “ACO RHCs” for which there were data to calculate relative disparities in diabetes-related hospitalization rates of Latino patients served by RHCs compared to Whites. Over the study period, the mean relative disparity declined from 0.148 in 2008 to −0.250 in 2015. That is, for 2008, the percentage difference in mean diabetes-related hospitalization rates for Latinos relative to Whites was +14.8%. For 2015, the percentage difference in mean rates of Latinos relative to Whites was −25.0%. As illustrated in Figure 3 when comparing the Post-ACO period (2013 – 2015) to the Pre-ACO period (2008 – 2011), there appears to be a decrease in relative disparities of mean diabetes-related hospitalization rates for Latinos.

Table 3.

Descriptive Statistics for Relative Disparities, “ACO RHCs,” CA, FL, TX: 2008 – 2015 (n = 116)

	2008	2009	2010	2011	2012	2013	2014	2015
Mean	.147766	−.074052	−.027803	−.053623	−.092707	−.239420	−.227054	−.250061
Median	.081477	−.095084	−.059089	−.098498	−.132691	−.259041	−.228844	−.285442
Standard Deviation	.604254	.234485	.254889	.288801	.366153	.281342	.194162	.279438

Note. Values were calculated for relative disparities (in diabetes-related hospitalizations) between older rural Latino and White Medicare beneficiaries in a panel of Rural Health Clinics participating in Medicare ACOs and located in California, Florida, and Texas, 2008 – 2015.

Figure 3. Trend in Relative Disparities, “ACO RHCs,” CA, FL, TX: 2008 – 2015 (n = 116).

Note. Trend line illustrates the mean relative disparities (in diabetes-related hospitalizations) between older rural Latino and White Medicare beneficiaries in a panel of Rural Health Clinics participating in Medicare ACOs and located in California, Florida, and Texas, 2008 – 2015.

Pre- and Post-ACO Years Comparison Results

In Table 4, we present findings of the comparison of the Pre-ACO period to the Post-ACO period for the initial nationwide panel of 2,683 RHCs throughout the U.S. The mean absolute disparity in diabetes-related hospitalization rates was – 0.004 for the Pre-ACO period (2008 – 2011), as compared to −0.011 for the Post-ACO period (2013 – 2015.) The mean absolute disparity for the Post-ACO period was found to be significantly smaller than that of the Pre-ACO period. This result may be interpreted as follows: the mean diabetes-related hospitalization rate for Latinos as compared to their White counterparts was lower during the Post-ACO period.

Table 4.

Test for Difference in Absolute Disparities, Pre- and Post-ACO Years, Nationwide Panel: 2008 – 2015 (n = 2,683)

Measures	Mean	Difference in Mean	95% Confidence Interval	P-Value
Absolute Disparity, Pre-ACO Years (2008 – 2011)	−0.004	-	−0.006 to −0.002	0.000
Absolute Disparity, Post-ACO Years (2013 – 2015)	−0.011	-	−0.012 to −0.009	0.000
Difference between Absolute Disparity Means, Pre- and Post-ACO Years	-	−0.006	−0.009 to −0.004	0.000

Discussion

This investigation analyzed multiple years of data for older rural Latino Medicare beneficiaries served by RHCs. The analyses enabled the researchers to measure health disparities for rural Latinos as related to diabetes. The results of this empirical investigation give insight into how diabetes-related health outcomes have changed over a period of time before and after implementation of the Medicare ACO model.

In this section we describe some of the broad findings resulting from this investigation. First, for the panel of RHCs throughout the U.S., there was a decrease in the mean disparities in diabetes-related hospitalization rates over the eight-year study period. The trend in mean diabetes-related disparities for the “ACO RHCs” in the three-state sample (for California, Florida and Texas) mirrored that of the nationwide panel. Thus, for both the nationwide panel and the “ACO RHC” sample, there was a decrease in the mean diabetes-related hospitalization rates for Latinos as compared their White counterparts.

Second, the comparison of a three-year time period after Medicare ACO implementation in 2012 to a four-year period before the implementation resulted in some general findings. For the nationwide panel, there was a statistically significant difference between the two time periods when comparing average disparities. That is, when comparing the Post-ACO period to the Pre-ACO period, the diabetes-related hospitalization rates of rural older Latinos were, on average, lower than the rates of their White counterparts.

The overall trends we found are consistent with other recent research findings. For example, the “Diabetes Report Card 2014” reported that there has been an increase in the number of diabetes-related hospitalizations in female, older, White and/or privately insured patients (CDC, 2014.)

It is important to note that because the nationwide panel included both “ACO RHCs” and “non-ACO RHCs,” the decrease in disparities may be related to factors in addition to, or apart from, ACO participation by RHCs. That is, a number of possible factors may contribute to the decrease diabetes-related hospitalization rates for Latinos in more recent years. An initial interpretation is that the health outcomes (as measured by hospitalizations) for Latinos improved over time. With their increase in numbers throughout the U.S., including in rural areas, Latinos may have greater exposure to health education opportunities that improve their lifestyles and health. Over time, they may have become more aware of diabetes prevention and management measures, and more likely to incorporate them into their daily activities. From the provider standpoint, the shift in the U.S. healthcare system from an emphasis on volume to one of value has placed greater emphasis on prevention and management of disease. Through a variety of alternative payment models, such as ACOs, healthcare providers are incented to improve healthcare quality through the provision of prevention and management services. Other quality improvement initiatives such as the CMS Hospital Readmissions Reduction Program may also contribute to declines in hospitalization rates in the more recent years.

Recent advances in technology have greatly improved the delivery of healthcare services in rural areas and may contribute to reducing health disparities among rural populations. Internet access in rural areas has been facilitated with bandwidth improvements. Improvements in technology have made it easier for rural healthcare providers to monitor patient compliance and follow up as needed.

Other interpretations of the decrease in Latino hospitalization rates merit exploration as well. For example, some Latino patients may not have a full understanding of the importance or urgency of recommended hospitalization because of language barriers or cultural differences. A number of studies suggest that better communication between Latino patients and their providers could improve their health care and health outcomes, and help to reduce Latino health disparities in the U.S. Language barriers contribute to Spanish-speaking Hispanics being less likely to receive preventive health services (DuBard & Gizlice, 2008; Hargraves, 2001.) In recent years, the increased availability of live video interpreter services has facilitated provider communication with non-English speakers. However, in addition to language, more subtle forms of communication such as culturally-specific differences in gestures and eye contact may detract from optimal communication between the provider and the Latino patient. On the other hand, culturally-specific values and lifestyle may diminish the receptiveness of the Latino patient to taking advice or instructions given by the provider. These more subtle forms of communication with Latino patients should improve over time as Latinos become more integrated into the U.S. population – in rural as well as urban areas.

It is important to note that although a statistically significant decrease in diabetes-related disparities between Latino and White patients was found, the changes from year-to-year were relatively small (less than one percent.) Many factors may have contributed to the decline in disparities. Future research will expand on the analysis of the contributors to this trend, including the possible influence of ACO participation by RHCs. In this investigation we did not analyze the unique contribution of Medicare ACO participation by RHCs on diabetes-related hospitalizations. Phase two of the larger study is designed to meet this need.

Despite the limitations, this investigation had several strengths. First, data for a large, nationwide sample of almost 75,000 Medicare beneficiaries served by RHCs were analyzed, including data for rural Latino as well as rural White patients. In addition, the data include the health outcomes of patients representing three states (California, Florida, and Texas) and three different regions of the U.S. (HHS Regions 9, 4 and 6 respectively.) The health disparities of all three states revealed similar trends overall during the study period, with some variation.

Conclusions

The purpose of this investigation was to quantify and describe any disparities between rural Latinos and their White counterparts using one measure of health outcomes: diabetes-related hospitalizations. In addition, we analyzed whether there was a difference in the mean disparities in the years before Medicare ACO implementation and after.

Latinos are a growing, multicultural sector of the U.S. population. Along with other groups, many Latinos experience factors that detract from good health and health outcomes, including language barriers and cultural differences, as well as poverty, lower educational attainment, and other unfavorable social determinants of health. For Latinos and others living in rural areas, geographic access to healthcare services compounds their vulnerability to adverse outcomes of acute illness and chronic conditions.

Future research will continue to examine trends in diabetes incidence and diabetes-related health outcomes. With the continued growth of the Latino population, it is critical that factors that contribute to their health and well-being be identified and promoted. The awareness of the importance of decreasing health disparities will advance the development of healthy communities throughout the U.S.