Federally Qualified Health Center Use Among Dual Eligibles: Rates Of Hospitalizations And Emergency Department Visits

Abstract

People who are eligible for both Medicare and Medicaid, known as “dual eligibles,” disproportionately are members of racial or ethnic minority groups. They face barriers accessing primary care, which in turn increase the risk of potentially preventable hospitalizations and emergency department (ED) visits for ambulatory care–sensitive conditions. Federally qualified health centers provide services known to address barriers to primary care. We analyzed 2008–10 Medicare data for elderly and nonelderly disabled dual eligibles residing in Primary Care Service Areas with nearby federally qualified health centers. Among our findings: There were fewer hospitalizations for ambulatory care–sensitive conditions among blacks and Hispanics who used these health centers than among their counterparts who did not use them (16 percent and 13 percent fewer, respectively). Use of the health centers was also associated with 3 percent and 12 percent fewer hospitalizations for ambulatory care–sensitive conditions among nonelderly disabled blacks and Hispanics, respectively. These findings suggest that federally qualified health centers can reduce disparities in preventable hospitalizations for some dual eligibles. However, further efforts are needed to reduce preventable ED visits among dual eligibles receiving care in the health centers.

Created in 1965 as part of President Lyndon Johnson’s War on Poverty, federally qualified health centers are federally funded facilities that are legally mandated to provide primary care. The health centers use an income-sensitive sliding scale of fees to serve all patients without regard for their ability to pay. In 2013 some 1,202 federally qualified health centers provided care for approximately 21.7 million patients.¹ The Affordable Care Act is expected to increase the health centers’ capacity. Data from 2012 indicate that nearly 93 percent of patients at the health centers have incomes below 200 percent of the federal poverty level, and approximately 69 percent are members of a racial or ethnic minority group.²

Most health center patients are uninsured or enrolled in Medicaid. However, the centers also served over 1.7 million Medicare beneficiaries in 2009—an increase of more than 165 percent since 1996.¹

Previous studies found that areas served by a federally qualified health center have both lower rates of emergency department (ED) use and lower rates of hospitalizations for ambulatory care–sensitive conditions.^3–9 These are conditions that can be prevented or effectively managed with regular use of primary care services and for which hospital-based care (hospitalizations and ED visits) is typically appropriate only when the condition has progressed in the absence of adequate primary care. For example, congestive heart failure, hypertension, and diabetes are all conditions that can be controlled in an outpatient setting with appropriate medications and proper care management.

However, no previous studies have evaluated the relationship between receiving care at a federally qualified health center and rates of hospital-based care for ambulatory care–sensitive conditions among people who are eligible for both Medicare and Medicaid, known as “dual eligibles.” Moreover, we know of no previous work that has described the characteristics of dual eligibles who use federally qualified health centers.

Dual eligibles are of particular concern to policy makers, health care providers, and others because they have substantial health care needs that are often unmet. This population also has increased morbidity and mortality and disproportionately high costs of care. For instance, rates of hospital-based care for some ambulatory care–sensitive conditions are twice as high among dual eligibles, compared to the rest of the Medicare population.¹⁰ In fact, among dual eligibles ages sixty-five and older, approximately 25 percent of hospitalizations are potentially preventable.¹¹

The overall objective of this study was to determine the association between use of federally qualified health centers and ambulatory care–sensitive hospital and ED visits among dual eligibles. Elderly blacks and Hispanics are six times as likely as elderly whites to be dual eligibles, and members of racial or ethnic minority groups are at greater risk than non-Hispanic whites for preventable hospital-based care.¹² Thus, we also determined whether use of the health centers was associated with racial or ethnic disparities in rates of ambulatory care–sensitive admissions and ED visits.

Given previous research, we hypothesized that dual eligibles who visited a federally qualified health center would have lower rates of hospital-based care for ambulatory care–sensitive conditions than dual eligibles who did not visit a health center. Additionally, because the centers appear to be effective at reducing racial disparities in care,^13,14 we hypothesized that the association between center use and lower rates of hospital-based care for ambulatory care–sensitive conditions would be stronger among black and Hispanic dual eligibles than among whites.

Study Data And Methods

SOURCES OF DATA AND STUDY POPULATION

We obtained 100 percent Medicare Part A inpatient and 100 percent Part B institutional outpatient claims and enrollment data for 2008–10. The Part A and Part B Medicare claims data contain service dates, diagnoses, services provided, billed charges, and approved reimbursements. Additionally, outpatient claims data specify the place of service. The enrollment data provided information about demographic characteristics such as age, sex, race/ethnicity, and place of residence. We linked these data together using personal identifiers.

As shown in online Appendix Exhibit 1, we constructed our sample on an annual basis.¹⁵ Starting with 28,660,177 person-year observations, we used the Medicare enrollment data to restrict our analysis to people who were enrolled in Medicare Part A, Medicare Part B, and Medicaid for an equal number of months. This left us with 22,850,440 person-year observations. In addition, we eliminated 4,753,416 person-year observations pertaining to dual eligibles with one or more months of Medicare Advantage enrollment, because no claims data areavailable for this population.

Using the outpatient claims data, we identified care provided by federally qualified health centers by the presence of type-of-bill code 731 or 771 on the claim (these values indicate that the facility filing the claim is a federally qualified health center). If a dual eligible visited a health center at least once during a calendar year, we considered that person to be a health center user for that year.

To ensure that health center users were compared only with nonusers who could reasonably access a center, we included only health center nonusers who resided in a Primary Care Service Area containing a center. The Dartmouth Atlas of Health Care developed the Primary Care Service Area designation, under contract to the National Center for Health Workforce Analysis, to “reflect Medicare patient travel to primary care providers.”¹⁶

We matched dual eligibles with their Primary Care Service Area following the methodology recommended by the Dartmouth Atlas.¹⁷ Under this process, each ZIP code is assigned to a ZIP code tabulation area for each year. Each ZIP code tabulation area is assigned to a Primary Care Service Area for the duration of the census period. For 2010 we used the 2009 assignment process, which remains the most recent available.

To identify people living in a Primary Care Service Area with a federally qualified health center, we matched all national provider identifiers listed on claims with type-of-bill code 731 or 771 with their corresponding Primary Care Service Area based on the ZIP code associated with the national provider identifiers. Dual eligibles residing in one of these Primary Care Service Areas and those who visited a federally qualified health center at any point during the year were retained for analysis. All others were excluded, as shown in Appendix Exhibit 1.¹⁵

Finally, we excluded people who did not reside in the District of Columbia or one of the fifty states, and people whose Primary Care Service Area or metropolitan area of residence could not be determined. Also excluded were people whose age was unrealistic (for example, dual eligibles who were younger than age sixty-four at the beginning of the year they became eligible for Medicare on the basis of age). Our final analytic sample consisted of 10,268,737 person-year observations.

VARIABLES

Our key outcome variables were hospitalizations and ED visits for ambulatory care–sensitive conditions. The Agency for Healthcare Research and Quality and others have identified conditions that can be prevented or effectively managed with regular use of primary care services. Rates of ED use and hospitalization for these conditions have been widely used as an indicator of limited access to primary care, with variation in these rates across racial or ethnic groups possibly reflecting racial or ethnic disparities in primary care access.^{3,6,8,10,18,19}

We identified hospitalizations for ambulatory care–sensitive conditions using all Part A claims submitted on behalf of the study population, and we identified ED visits using Part B and Part A claims listing any revenue code 0450–0459 (ED) or 0981 (professional fees—ED). Visits to the ED resulting in hospitalizations were defined as both ED visits and hospitalizations. Since we conducted analyses separately for each outcome, it was appropriate to include these events in both analyses. We used Prevention Quality Indicator software available from the Agency for Healthcare Research and Quality to flag hospitalizations and ED visits for ambulatory care–sensitive conditions.²⁰

ANALYSES

We compared the characteristics of dual eligibles who used federally qualified health centers with those of nonusers. Geography may play an important role in patients’ decisions about using the health centers. Therefore, we calculated the distance from the centroid—the geographic center—of the patient’s ZIP code to the centroid of the ZIP code of the nearest health center within the same Primary Care Service Area. We confirmed our findings with an unbalanced analysis of variation to assess whether the average distance differed after effects at the level of the Primary Care Service Area were controlled for.

Next, we calculated national rates of hospital-based care for ambulatory care–sensitive conditions among dual eligibles by reason for eligibility, status as a user or a nonuser of federally qualified health centers, and race or ethnicity.

Finally, we constructed a series of adjusted relative rates to analyze disparities in hospital and ED utilization for ambulatory care–sensitive conditions. First, we directly standardized rates by age, sex, census region, and metropolitan area of residence (using the full sample of dual eligibles as the standard population). This direct standardization adjusted for differences in the demographic composition of population subgroups. We then divided the standardized rate of ambulatory care–sensitive condition hospitalizations and ED visits for each nonwhite racial or ethnic group of federally qualified health center users by the standardized rate of the hospitalizations and ED visits for white health center users to obtain adjusted relative rates, which could be used to examine the extent of racial or ethnic disparities. We repeated these calculations for nonusers.

Differences in relative rates across racial or ethnic groups can be interpreted as population-level disparities in hospital-based care for ambulatory care–sensitive conditions. For the relative rates, we used two-sample t-tests of significance with an alpha value of 0.05, and we report 95 percent confidence intervals.

This study protocol was approved by the Institutional Review Boards of the University of Iowa and Brown University. Use of the data was authorized by the Centers for Medicare and Medicaid Services.

LIMITATIONS

Our study was subject to several limitations. First, the data for this study were limited to information about dual eligibles enrolled in fee-for-service Medicare. Therefore, the results of this study are not necessarily generalizable to the Medicare managed care population.

Second, this study did not control for other sources of primary care received by dual eligibles. As a result, we were unable to reach any conclusions about the use of federally qualified health centers as a usual source of care.

Third, our data did not allow us to identify whether or not someone was residing in an institution. It has been established that institutionalized beneficiaries have rates of potentially preventable hospitalization that are nearly ten times higher than those of the community-based population.²¹ Thus, if receipt of care at a federally qualified health center is negatively associated with residing in an institution, this might introduce some bias to our results. However, one recent report indicated that 87 percent of dual eligibles are community based, so the scope of this potential bias is likely to be limited.²²

Fourth, there might be self-selection among users of federally qualified health centers that we were unable to control for with our data.

Finally, the Medicare claims data contained information on diagnoses that are generally considered potentially preventable causes of hospitalization and ED visits. Nonetheless, we lacked information on the severity of illness and other factors, which makes it possible that some hospitalizations and ED visits were not actually preventable.

Study Results

In our sample of dual eligibles living in Primary Care Service Areas with a federally qualified health center, there were nearly 5.3 million hospitalizations and 11.8 million ED visits during the three-year study period. Of these, more than 1.0 million hospitalizations (19.2 percent) and nearly 1.7 million ED visits (14.2 percent) were classified as potentially preventable. The number of dual eligibles receiving care at a health center increased from 446,329 in 2008 to 542,777 in 2010. In 2008, 44.9 percent of dual-eligible health center users were elderly, 54.1 percent were nonelderly disabled, and 1.1 percent had end-stage renal disease. In 2010, the percentages were 42.7 percent, 56.1 percent, and 1.2 percent, respectively.

DIFFERENCES BETWEEN HEALTH CENTER USERS AND NONUSERS

The characteristics of our sample population, stratified by federally qualified health center use, are shown in Exhibit 1; standard deviations are reported in Appendix Exhibit 2.¹⁵ Users were, on average, about five years younger than nonusers, about 12 percentage points less likely to live in a metropolitan area, and less likely to live in the South or Midwest, compared to nonusers. Also compared to non-users, users resided slightly closer to the location of health centers. Other differences (such as race or ethnicity and hospitalization rates) were minimal. However, all differences were significant. Among federally qualified health center users, 18.8 percent of hospitalizations and 12.0 percent of ED visits were potentially preventable, compared to 19.2 percent of hospitalizations and 14.7 percent of ED visits among nonusers (data not shown).

EXHIBIT 1.

Characteristics Of Dual-Eligible Medicare Enrollee Users And Geographically Matched Nonusers Of Federally Qualified Health Centers (FQHCs)

Characteristic	Users	Nonusers
Mean age (years)	59.7	64.6
Male (%)	39.8	38.8
Metropolitan (%)	73.6	85.2
Race/ethnicity (%)
White	60.3	60.1
Black	23.4	23.1
Asian	4.0	5.4
Hispanic	9.0	8.0
Other	3.4	3.3
Region (%)
Northeast	19.9	16.3
Midwest	17.4	21.1
South	33.9	38.1
West	28.9	24.6
Mean distance from residence to FQHC (miles)	2.4	3.7
Mean annual number of:
FQHC visits	5.1	0.0
Hospitalizations	0.5	0.5
ACS hospitalizations	0.1	0.1
ED visits	1.5	1.1
ACS ED visits	0.2	0.2
Current reason for Medicare eligibility (%)
Elderly	43.7	56.4
Nonelderly disabled	55.2	41.8
End-stage renal disease	1.1	1.8

SOURCE Authors’ analysis of Medicare claims data for 2008–10. NOTES All differences between FQHC users (n = 1, 475, 617 person-years) and nonusers (n = 8, 793, 120 person-years) were significant (p < 0.01), as determined by two-sample t-tests for continuous variables and Pearson chi-square tests for categorical variables. ACS is ambulatory care sensitive. ED is emergency department.

UNADJUSTED RATES OF HOSPITAL-BASED CARE

Both hospitalization and ED visit rates for ambulatory care–sensitive conditions were lower among the nonelderly disabled population than the elderly population (Exhibit 2). Among both the elderly and nonelderly disabled populations, hospitalization rates for such conditions were lower among users of federally qualified health centers than among nonusers, but ED visit rates were higher among users than among nonusers. This pattern largely held within racial and ethnic groups as well. Both hospitalization and ED visit rates for ambulatory care–sensitive conditions were highest for blacks and second-highest for whites, in both the elderly and nonelderly disabled populations.

EXHIBIT 2.

Unadjusted Annual Rates Of Ambulatory Care–Sensitive Hospital Events Among Elderly And Nonelderly Disabled Dual-Eligible Medicare Enrollees, Per 1,000 Person-Years

	Elderly				Nonelderly disabled
	Hospitalizations		ED visits		Hospitalizations		ED visits
	Number	SD	Number	SD	Number	SD	Number	SD
STUDY POPULATION
All	120.6	460.5	173.4	591.2	65.9	391.3	138.8	655.3
White	128.2	469.1	189.2	600.3	62.5	366.7	132.4	608.3
Black	145.1	521.4	214.4	692.2	81.6	464.5	169.4	781.0
Asian	49.0	276.9	68.3	334.7	30.5	264.4	68.2	705.5
Hispanic	96.8	414.8	138.0	502.8	47.8	335.6	104.0	533.1
Other	88.6	390.9	113.7	471.8	53.2	333.8	108.6	560.4
HEALTH CENTER USERS
All	106.7	440.4	189.2	654.8	64.9	393.3	169.7	741.5
White	119.1	465.0	208.2	677.0	62.3	369.9	166.6	704.2
Black	116.6	466.9	211.8	737.9	80.1	473.7	198.8	881.3
Asian	42.1	265.4	73.7	369.8	34.7	283.9	89.5	479.9
Hispanic	74.3	356.6	138.8	516.6	42.2	292.3	117.5	526.7
Other	88.7	390.4	153.6	601.5	56.8	365.1	149.3	725.1
HEALTH CENTER NONUSERS
All	122.4	463.0	171.4	582.4	66.1	390.8	131.9	634.4
White	129.4	469.6	180.8	589.9	62.6	366.0	124.8	584.5
Black	149.0	528.3	214.8	685.7	81.9	462.5	163.2	758.1
Asian	49.7	278.2	67.7	330.5	29.7	260.4	64.1	741.6
Hispanic	100.3	423.1	137.9	500.6	49.3	345.9	100.6	534.7
Other	88.6	391.0	108.5	451.7	52.3	325.6	98.5	511.4

SOURCE Authors’ analysis of Medicare claims data for 2008–10. NOTES ED is emergency department. SD is standard deviation.

ADJUSTED RELATIVE RATES FOR RACIAL/ETHNIC GROUPS

Federally qualified health center use was associated with a reduction in hospitalizations for ambulatory care–sensitive conditions among elderly black and Hispanic Medicare beneficiaries but was associated with an increase in the rate among all other racial or ethnic groups (Exhibit 3). Specifically, the rates were 16 percent lower for blacks and 13 percent lower for Hispanics. Asians experienced the greatest increase: 13 percent.

EXHIBIT 3.

Adjusted Relative Rates Of Hospital-Based Care For Ambulatory Care–Sensitive Conditions Among Elderly And Nonelderly Disabled Dual-Eligible Medicare Enrollee Users And Nonusers Of Federally Qualified Health Centers, By Race/Ethnicity

Race/ethnicity	Elderly				Nonelderly disabled
	Hospitalization		ED visit		Hospitalization		ED visit
	Rate	95% CI	Rate	95% CI	Rate	95% CI	Rate	95% CI
White	1.01	1.01, 1.01	1.20	1.19, 1.20	1.04	1.04, 1.04	1.36	1.35, 1.36
Black	0.84	0.84, 0.85	1.03	1.02, 1.03	0.97	0.97, 0.98	1.23	1.22, 1.23
Hispanic	0.87	0.86, 0.87	1.10	1.09, 1.11	0.88	0.88, 0.89	1.08	1.07, 1.09
Asian	1.13	1.12, 1.14	1.27	1.26, 1.28	1.02	1.00, 1.04	1.09	1.07, 1.12
Other	1.06	1.05, 1.08	1.36	1.35, 1.38	1.00	0.99, 1.02	1.37	1.36, 1.39

SOURCE Authors’ analysis of Medicare claims data for 2008–10. NOTES The relative rates indicate differences between users and nonusers of federally qualified health centers within racial/ethnic groups. A rate greater than 1.0 indicates that use of the centers is associated with an increased rate of ambulatory care–sensitive events (hospitalizations or emergency department [ED] visits). A rate of less than 1.0 indicates that use of the centers is associated with a decreased rate of the events. For instance, the elderly black rate of 0.84 for hospitalizations means that elderly black center users had 16 percent fewer ambulatory care–sensitive events than elderly black center nonusers. Rates are adjusted for age, sex, census region, and metropolitan area of residence. CI is confidence interval.

Use of the health centers was not associated with a reduction in hospitalizations for ambulatory care–sensitive conditions among black non-elderly disabled Medicare beneficiaries to the same extent as it was among black elderly beneficiaries (Exhibit 3). Among nonelderly disabled Medicare beneficiaries who used a federally qualified health center, there was a 3 percent reduction in hospitalizations for ambulatory care–sensitive conditions for blacks. In this group, whites experienced the greatest increase: 4 percent.

Federally qualified health center use also appeared to be associated with a substantial increase in the rate of ED visits for ambulatory care–sensitive conditions among both groups of Medicare beneficiaries, with relative rates ranging from a 3 percent increase among elderly black beneficiaries to a 37 percent increase for nonelderly disabled beneficiaries in the racial/ethnic category of “other.”

ADJUSTED RELATIVE RATES BETWEEN RACIAL/ETHNIC GROUPS

Exhibit 4 presents the relative rates of hospitalizations and ED visits for ambulatory care–sensitive conditions between members of racial or ethnic minority groups and whites, looking separately at users and nonusers of federally qualified health centers. This allowed us to ascertain the extent to which use of the health centers might reduce racial or ethnic disparities in potentially preventable hospitalizations and ED visits.

EXHIBIT 4.

Adjusted Relative Rates Of Hospital-Based Care For Ambulatory Care–Sensitive Conditions Among Black, Hispanic, And Asian Dual-Eligible Medicare Enrollees Compared To White Dual Eligibles, By Use Of Federally Qualified Health Centers (FQHCs)

Race/ethnicity	Elderly				Nonelderly disabled
	Hospitalization		ED visit		Hospitalization		ED visit
	Rate	95% CI	Rate	95% CI	Rate	95% CI	Rate	95% CI
HEALTH CENTER USERS
Black	0.96	0.96, 0.97	1.01	1.01, 1.02	1.22	1.21, 1.22	1.16	1.15, 1.16
Hispanic	0.72	0.72, 0.73	0.76	0.75, 0.76	0.76	0.75, 0.77	0.72	0.71, 0.72
Asian	0.54	0.54, 0.55	0.49	0.49, 0.50	0.57	0.56, 0.58	0.52	0.51, 0.53
Other	0.82	0.81, 0.83	0.82	0.81, 0.83	0.90	0.89, 0.91	0.90	0.89, 0.91
HEALTH CENTER NONUSERS
Black	1.16	1.15, 1.16	1.18	1.18, 1.18	1.30	1.29, 1.30	1.28	1.28, 1.29
Hispanic	0.84	0.84, 0.84	0.82	0.82, 0.83	0.89	0.89, 0.90	0.90	0.90, 0.91
Asian	0.49	0.48, 0.49	0.46	0.46, 0.46	0.58	0.58, 0.59	0.64	0.64, 0.65
Other	0.78	0.78, 0.78	0.72	0.72, 0.72	0.93	0.92, 0.93	0.89	0.88, 0.89

SOURCE Authors’ analysis of Medicare claims data for 2008–10. NOTES The relative rates indicate differences between the racial/ethnic groups and whites by FQHC use. A rate greater than 1.0 indicates that the racial/ethnic group has an increased rate of ambulatory care–sensitive events (hospitalizations or emergency department [ED] visits) compared to whites. A rate of less than 1.0 indicates that the racial/ethnic group has a decreased rate of the events compared to whites. A lower rate among FQHC users versus nonusers for a given racial/ethnic group indicates a reduction in disparities. Rates are adjusted for age, sex, census region, and metropolitan area of residence. ED is emergency department. CI is confidence interval.

For instance, we found that elderly black non-users were hospitalized for an ambulatory care–sensitive condition 16 percent more often than elderly white nonusers. In contrast, elderly black health center users were hospitalized for such a condition 4 percent less often than elderly white health center users. This finding indicates that use of federally qualified health centers has the potential to reduce racial or ethnic disparities in hospitalizations for ambulatory care–sensitive conditions among this segment of the dual-eligible population. Our results demonstrate that racial and ethnic disparities in hospital-based care for ambulatory care–sensitive conditions vary widely according to Medicare eligibility category, the outcome measure in question, and use of federally qualified health centers.

For elderly Hispanic dual eligibles, we observed a similar reduction, albeit smaller in magnitude (Exhibit 4). Among users and nonusers of the health centers, Asians and people in the racial/ethnic category of “other” were less likely than whites to be hospitalized or have an ED visit for an ambulatory care–sensitive condition.

For all races or ethnicities in the nonelderly disabled population, use of federally qualified health centers—except for ED visits among the racial/ethnic category of “other”—was associated with a decrease in the rate of hospitalizations and ED visits for ambulatory care–sensitive conditions relative to whites. This suggests that health center use has the potential to reduce disparities for these groups of Medicare beneficiaries. Among nonelderly disabled people, we observed significantly higher rates of both hospitalizations and ED visits for ambulatory care–sensitive conditions for blacks relative to whites. However, the relative disparity was attenuated among health center users.

Discussion

In this study we sought to determine the association between use of federally qualified health centers and hospitalizations and ED visits for ambulatory care–sensitive conditions among dual eligibles. We were particularly interested in determining whether health center use by this population attenuated racial or ethnic disparities in the rates of hospitalizations and ED visits for these conditions. High rates of hospitalizations and ED visits for the conditions—and racial or ethnic disparities in these rates—are of concern because they are indicative of inadequate access to primary care.

Generally, we found that use of federally qualified health centers was associated with lower rates of hospitalization for ambulatory care–sensitive conditions among black and Hispanic dual eligibles, regardless of their reason for Medicare eligibility, but the relationship was considerably stronger among elderly dual eligibles. By contrast, health center use was associated with higher rates of hospitalization for ambulatory care–sensitive conditions among whites, Asians, and other races (Exhibit 3).

We also found that ED visit rates for ambulatory care–sensitive conditions were higher among users of federally qualified health centers, compared to nonusers, for both elderly and nonelderly disabled dual eligibles. Given that inpatient admissions often occur via the ED, this is an interesting set of results. It suggests that although health center users, particularly blacks and Hispanics, are less likely to be hospitalized for an ambulatory care–sensitive condition, they receive a larger proportion of care for these conditions via ED visits. This finding should prompt further study of the causes of preventable ED visits among dual eligibles who receive care in federally qualified health centers and of ways to reduce these visits.

Relative rates of hospital-based care for ambulatory care–sensitive conditions across racial and ethnic groups indicated that among elderly Medicare beneficiaries, use of federally qualified health centers has the potential to reduce disparities in the rates of hospitalizations and ED visits for ambulatory care–sensitive conditions for blacks and Hispanics compared to whites. Most strikingly, health center use was associated with the elimination of—or even a slight reversal in—the black-white disparity in hospitalizations for ambulatory care–sensitive conditions among elderly Medicare beneficiaries. Among non-elderly disabled Medicare beneficiaries, blacks who used a health center had lower rates of hospitalization and ED visits for ambulatory care–sensitive conditions relative to nonusers, despite having significantly higher rates of hospitalization and ED visits for these conditions relative to whites.

Racial disparities in potentially preventable hospitalizations have been well documented.²³ Such hospitalizations are more frequent among beneficiaries who are elderly, those who are black, and those who live in either rural or metropolitan areas (that is, not in suburban areas).¹⁹ They are also more common among people with low socioeconomic status,²⁴ including the uninsured and Medicaid beneficiaries,²⁵ and among people living in areas with a lower per capita supply of primary care physicians.²⁶

Groups with disproportionately high rates of potentially preventable hospitalizations also have disproportionately high rates of potentially preventable ED visits, even after disease prevalence and severity are controlled for. Inadequate primary care has been suggested as an explanatory factor.²⁷

Hospital-based care for ambulatory care–sensitive conditions among the dual-eligible population likely reflects inadequate access to primary care, and it represents an attractive target for quality improvement and cost savings in Medicare and Medicaid. Our study provides moderate evidence to support the use of federally qualified health centers in reducing potentially preventable hospital-based care among certain groups of Medicare beneficiaries. However, it is important to consider how such a strategy might be implemented, beginning with understanding patterns of health center use among dual eligibles.

During our three-year study period, there was an upward trend in the number of dual eligibles who visited a federally qualified health center for care. Our analysis indicates that more than a half-million dual eligibles used the health centers in 2010. However, this figure represents just 8.6 percent of the fee-for-service dual-eligible population, which suggests that there is substantial potential to increase health center use in this population.

In part, the low rate of health center use by dual eligibles is because, by our analysis, 41.2 percent of fee-for-service dual eligibles do not live near a center. It is unclear from the current study to what extent dual eligibles reside in areas that would be capable of sustaining a health center if one were established there. However, policy makers should consider obtaining information on where large numbers of dual eligibles without access to a health center live, and evaluating whether future expansions of the federally qualified health center program could be effectively targeted to these areas to serve more dual eligibles.

Still, our data suggest that roughly 85 percent of fee-for-service dual eligibles living near a federally qualified health center do not use the health center, despite its availability. This highlights the need for further research into the reasons why a sizable majority of dual eligibles with geographic access to a health center opt never to visit it for care. Our results suggest that the distance between a dual eligible’s home and the nearest health center may partially explain usage patterns.

Other explanations include the perceived stigma of receiving care from a safety-net provider, having already established another usual source of care, or residing in a long-term care facility. Federally qualified health centers do provide primary care to residents of long-term care facilities. However, the extent to which this occurs among dual eligibles is unknown and warrants further study. Health center users are also significantly less likely to reside in metropolitan instead of nonmetropolitan areas and may visit health centers for care because there are fewer alternatives for care in rural areas.

Finally, given the finding that use of federally qualified health centers is associated with increased use of the ED for ambulatory care–sensitive conditions, policy makers should continue to consider implementing programs to reduce preventable ED visits. This may mean enhancing communication between EDs and federally qualified health centers and identifying ways for centers to keep their established patients from visiting the ED for nonurgent reasons.

The Center for Medicare and Medicaid Innovation implemented a three-year Medicare federally qualified health center advanced primary care practice demonstration project that ended in October 2014.²⁸ Evaluation of this demonstration project may prove informative in developing a model of care coordination that leads to reductions in inappropriate ED use.

Conclusion

This study provides the first empirical evidence of the association between use of federally qualified health centers and lower rates of hospitalizations for ambulatory care–sensitive conditions among black and Hispanic dual-eligible populations. Our results suggest that the health centers have the potential to reduce disparities in access to primary care among dual eligibles, which could improve health outcomes and reduce health care spending for this vulnerable population. Future research and policy efforts should focus on the role of federally qualified health centers as a usual source of care among dual eligibles.