In their role within the health care safety net, emergency departments (EDs) provide an important service to the public through 24-hour per day operations and identification and stabilization of patients with acute, life threatening conditions [1]. ED use is especially frequent among publicly insured and low-income populations [2]–[4], who disproportionately face worries about costs and coverage limitations and non-financial barriers such as lack of personal transportation and child care [4, 5]. Problems with access to primary care and concerns about costs, even for those who have health insurance, may be associated with more ED visits and poorer health for patients who visit the ED [6]. For public insurance programs, barriers that persist even in the face of health coverage represent a key target for improved services and outreach. This is especially important considering that a decreasing number of EDs are experiencing an increasing number of visits, and that ED visit rates for those covered by Medicaid account for much of the overall increase [7, 8].

Expansion of health coverage (e.g., under the Affordable Care Act) might help limit high ED utilization by improving insured access to primary care and other providers [9]. However, coverage is not equivalent to access [10]. Rather, an entire array of non-coverage related barriers to access—including office hours/availability, appointment problems, wait times, transportation, and lack of a usual primary care provider—likely drive ED visits and reliance on ED as a proxy source of primary care [11]–[15]. Moreover, given common perceptions of the ED as free or less expensive than regular care [2, 13] patients’ concerns about costs and coverage—which do not necessarily align with actual coverage status [16]—may also contribute to ED utilization patterns [17].

To better understand how financial and non-financial factors may be associated with frequent ED utilization among publicly insured patients, the present study assesses the associations that financial concerns about healthcare costs and coverage and practical/non-financial access barriers to care have with publicly insured individuals’ identification of the ED as usual source of care, and the actual number of ED visits. Measuring these associations increases our understanding of what may and may not work in seeking to address shortcomings in outreach, availability, and other persistent challenges that may be associated with reliance on, and frequent use of, ED services among publicly insured individuals.

Using administrative and survey data from a diverse sample of publicly insured adults in Minnesota, we analyzed the associations that financial concerns and non-financial access barriers had with identification of ED as usual source of care and actual ED visits. Financial concerns and non-financial barriers had different associations: after controlling for each other and covariates, robust associations remained mainly between: 1) financial concerns and identification of ED as usual source of care; and 2) non-financial access barriers and actual ED visits (from administrative data). The implications of this study are that policies which address financial concerns may decrease the already-low percentage of the publicly insured population that reports the ED as their usual source of care, but that limiting frequent actual ED use requires addressing the non-financial barriers that complicate non-emergent outpatient care.

Self-reported concerns and barriers among enrollees of public insurance programs, while previously identified as contributing to patterns of ED use [13, 14, 17], are actually associated with different challenges to reaching and accessing health care. On the one hand, financial concerns that exist despite Minnesota offering relatively generous coverage (compared to other states) [33] suggest a potential lack of awareness and knowledge of available services and costs. These financial concerns also may represent a lack of faith or trust that regular care is feasible—a belief that, in the current findings, coincides robustly with individuals’ belief that the ED is the go-to source for care. In an alternative formulation, it could be argued that concerns about coverage and identification of ED as usual source of care capture similar latent phenomena among publicly insured individuals’ worldview—an anxiety or distrust regarding regular services and reliance on putatively free ED services. Financial concerns also may reflect an actual awareness of specific financial disincentives for non-emergency use of the ED that currently exist in Minnesota’s public health care programs (copayments currently range from $3.50 upwards, depending on the program in which they are enrolled), although this is speculative and we are not aware of the extent to which the definition of emergency medical conditions is enforced.

In comparison, non-financial access barriers represent individual reports of experienced difficulty, often due to logistical, time- or resource-related constraints. While their relationship with reporting ED as usual source of care was statistically mediated by financial concerns, these non-financial access barriers were strongly associated with actual ED visits—similar to previous literature [13, 14], but here even controlling for individuals’ naming of ED as usual source of care, the presence of a mental health and injury or poisoning diagnosis, and office visit and inpatient stay utilization. This suggests that, rather than being associated with ED use simply as a correlate or expression of financial vulnerability, or indirectly as associated with poorer mental health, trauma, or broader utilization patterns, such non-financial barriers indicate practical difficulties that are embedded in the tangible constraints embedded in individuals’ everyday lives (e.g., child care, transportation, office hours, and others). It also suggests that enrollees’ financial concerns do not associate directly with ED use.

Together, findings suggest that coverage expansion (under PPACA or similar programs) will fail to address frequent ED use unless such expansion is paired some key efforts. These include outreach and education in helping enrollees understand what is available to them in non-ED care. Lack of knowledge about available services is a key concern for Minnesota’s publicly insured population, particularly those facing language and cultural barriers [34]. Unfortunately, county governments in Minnesota appear to shoulder much of the burden of outreach, but their financial capacity for outreach is already limited [35, 36], and it appears new PCACA provisions dedicated to improving health literacy may be poorly funded [37]. Moreover, as noted above, financial concerns are only associated with reporting the ED as usual source of care, but not actual ED visits. The rate of reporting the ED as usual provider is low (1.5%, which is comparable to previous studies finding between .9% and 4.6%, depending on the population [38, 39]). As such, increased awareness of what is covered or not covered may be indirectly related to ED visits (via identification of the ED as the usual provider). More directly, the key issues associated with actual ED visits, namely non-financial barriers, require more pointed efforts, considering that they pertain to accessibility of services (e.g., office hours) and limited everyday resources in one’s personal life (e.g., transportation or child care). As such, helping individuals identify a non-ED usual source of care is likely to be, at best, a half-measure, and at worst, may exacerbate some of the accessibility issues if coverage is expanded under the affordable care act—in other words, without better accessibility of services or consideration of other resources needed in patients’ lives, linking more publicly insured individuals with non-ED providers may be futile or simply stretch those providers’ capacity similar to what has happened with EDs. Addressing non-financial barriers requires finding ways to better target services with an eye toward the difficulties people face which are not cost/coverage-related. ED users often rely on hospital outpatient departments, community health centers, and other public health care clinics as a usual source of care [40]. Addressing non-financial barriers to these health care venues—such as expanding after-hours care, improving options for transportation, child care, and family responsibilities, or providing more information about available providers—may represent one strategy for limiting frequent ED use. Without outreach and supportive services, and potentially greater capacity of non-ED providers, coverage may not mean improvements in patterns of ED use.

The findings of this study should be read in light of its limitations. The data are secondary and cross-sectional, and so we are not able to truly test causality. Also, the sample, while diverse, cannot generalize to all publicly insured individuals in the U.S. However, Minnesota provides relatively comprehensive and affordable (e.g., low co-pay) coverage [33], and so findings here may represent a best-case scenario of the challenges that publicly insured individuals encounter. In a separate vein, although administrative data contained demographics and utilization, information on individuals’ incomes, health behaviors, or specific reasons for ED visits were unavailable. In addition, administratively derived variables may be subject to bias due to the fact that, because of sample size concerns, we did not limit analyses based on months of enrollment (which means that estimates may be affected by changes or lapses in enrollment). Sensitivity analyses of ED visit models based on consecutive enrollment showed that non-financial barriers, diagnosis indicators, and utilization measures remained significant, but the potential for bias should not be ignored. Certain survey measures here may be subject to recall bias, or to social desirability bias (e.g., underreporting of the ED as usual provider, or lower reporting of barriers as a problem, if respondents saw these as socially undesirable). Finally, the overall response rate (44%) was moderately low, but not surprising: response rates vary between 12% and 82% [41, 42], depending on follow-up efforts, and are often low among Medicaid and low-income populations—e.g., 38% [43] up to, in some cases, 50% [44]. Response rates here varied across adult sampling strata, ranging between 34.7% and 53% [18], and so we controlled for race and other demographics, along with weighting to control for sampling design. However, response rates remain a caveat, as with any study using survey data.

In sum, financial concerns were solely associated with reporting ED as usual provider, whereas non-financial access barriers were associated with actual ED use. Expanded coverage will not guarantee reductions in frequent ED use. Outreach/literacy efforts may help prevent viewing ED as one’s usual provider; improved availability of non-emergent and publicly covered services, particularly including supportive services such as transportation and child care, may help curb frequent ED visits.