Introduction

The Kaiser Family Foundation writes: “the impact of the ACA will depend on take-up of coverage among the eligible uninsured, and outreach and enrollment efforts will be an important factor in decreasing the uninsured rate” [1]. Health insurance is an important health determinant [2]; however, until recently one in five non-elderly adults in the United States was uninsured [3]. One goal of the Patient Protection and Affordable Care Act (ACA) is to increase health insurance enrollment in the United States [1], although its provisions do not offer universal access to coverage. In January 2014, the two major coverage provisions of the ACA were implemented: the creation of the Health Insurance Marketplace (HIM) with financial assistance to individuals and families with incomes below 400 % of the federal poverty line (FPL) and an expansion of Medicaid to individuals or families with incomes up to 138 % FPL in states that adopted Medicaid expansion. In the year after implementation, the rate of uninsured adults ages 18–64 fell from 20.4 to 16.3 % nationally [3]. In states that chose to expand Medicaid, the uninsured rate fell even further from 18.4 % in 2013 to 13.3 % in 2014 [3]. In mid-2015 the uninsured rate in Michigan, which expanded Medicaid in 2014, was 8.5 % [4].

Despite these gains, many Americans who now qualify for insurance programs and payment support under the ACA remain uninsured. A 2015 nationwide survey found that the majority of uninsured Americans recognized that health insurance is important and many were looking or planning to look for insurance [5]. However, almost half had not accessed the HIM and faced significant barriers to finding insurance, such as inadequate knowledge and the inability to navigate the insurance landscape. Reasons for remaining uninsured under the ACA are likely to vary by state and county, creating a need for local investigations to offer insights to national trends. A recent study in Utah, a state that did not expand Medicaid, demonstrated continued demand for care by uninsured patients [6]. These patients identified difficulty obtaining information, lack of instruction on how to apply, and the perceived expense of plans as primary barriers to accessing insurance.

Uniquely, Michigan implemented Medicaid expansion with the support of a Republican-dominated state legislature and a Republican governor, termed the Healthy Michigan Plan. The Healthy Michigan Plan included provisions requiring demonstration of statewide cost-savings within 3 years for continuation of the program [7]. Prior to the implementation of Medicaid expansion in Michigan, few low-income adults qualified for the program, though childless persons with incomes ≤35 % FPL were eligible for Medicaid benefits under the Adult Benefit Waiver [8, 9]. Following enactment of the ACA, an estimated 71 % of the 1.1 million previously uninsured Michigan residents became eligible for either the Healthy Michigan Plan or HIM tax credits [10]. The attention surrounding the ACA was estimated to enroll an additional 33,000 Michigan residents who had been eligible for traditional Medicaid but previously unenrolled [11]. Estimates from the Centers for Medicare and Medicaid Services (CMS) in 2015 indicated that Michigan Medicaid enrollment had increased by close to 600,000 since the expansion opened in April 2014 [12].

In spite of these gains, Michigan continues to have uninsured adults in need of health care. The University of Michigan Student-Run Free Clinic (UMSRFC) is a student organization through the University of Michigan Medical School dedicated to providing high quality health care to uninsured adults while creating opportunities for the university’s students and faculty physicians to address health disparities through direct action. The UMSRFC is open one half-day per week in Pinckney, MI, a town in the primarily rural Livingston County (population: 184,392). The clinic is the sole provider of free primary health care in Livingston County. Since the enactment of the ACA, states with expanded Medicaid have seen a decline in free clinic usage [13]; Livingston County saw the other free clinic provider close its doors in early 2015 due to decreased patient volume. However, UMSRFC still experiences health care demand from the uninsured in southeastern Michigan, providing care to approximately 400 unique patients each year.

In the changing insurance landscape, UMSRFC offers a unique opportunity to look at a cohort of Michigan residents who have not yet been able to benefit from the expansion of Medicaid and the establishment of the HIM. This study attempts to contribute local insights to the national dialogue about insurance by describing patients at UMSRFC: who they are, their activity in seeking health insurance, and their major barriers to becoming insured in a state with expanded Medicaid coverage.

Methods

Survey Administration

Over the 3-month period from March 14, 2015 to June 6, 2015, all UMSRFC patients were approached during their appointments to complete a questionnaire. At UMSRFC, patients are first seen by a team of one pre-clinical (year one or two) and one clinical (year three or four) medical students and then by a volunteer University of Michigan Medical School faculty physician. Patients were offered the survey privately, in an exam room during their clinic visits. Verbal consent was obtained, informing patients that they could choose not to answer any or all questions. Patients also signed a Notice of Privacy form upon initiating care with UMSRFC that included a clause consenting to research using de-identified data. University of Michigan undergraduate students, under the supervision of clinic managers and the researchers, administered the survey verbally to eliminate any effects from differing literacy levels. For patients who did not speak English, most commonly Spanish-speakers, trained University of Michigan medical student interpreters helped to administer the surveys verbally. The study was reviewed and approved by the University of Michigan Institutional Review Board (IRB).

Survey Content

The survey on health insurance consisted of seven questions concerning patients’ eligibility for and activity in pursuing health insurance options as part of a larger 20-question quality improvement questionnaire. The questions relating to this study were written by the authors based on conversations about insurance enrollment with patients, volunteers and Patient Financial Counselors at the clinic. The survey administration method had been tested and was ongoing for several months prior to the addition of the questions used in this study. Patients were asked about the size of their household, including spouses and dependents, and their estimated annual household income. These criteria were used to calculate the percentage of the Federal Poverty Level (FPL) for each patient to determine their eligibility for Medicaid and tax credits on the Health Insurance Marketplace (US residency or immigration status was not used in the assessment). The researchers used FPL to estimate Medicaid eligibility. Although Michigan has other requirements for adults to be eligible, the primary determinant is % FPL; most adults under 138 % FPL are eligible for the program.

Patients were asked about their activity in seeking health insurance. All patients were asked in binary yes/no questions: (a) “have you applied for Medicaid (or a Healthy Michigan Plan) in the last 6 months?” and (b) “have you looked elsewhere for health insurance, including healthcare.gov, in the last 6 months?” The 6-month interval was chosen to encompass the majority of the most recent HIM open enrollment period, from November 15, 2014 to February 15, 2015.

Patients who responded affirmatively to either question were defined as actively seeking insurance and were asked to describe the outcome as “accepted,” “denied,” or “in the process” for Medicaid or “able to find an acceptable plan,” “unable to find an acceptable plan,” or “in the process” for the Health Insurance Marketplace. Patients who responded negatively to both questions were defined as not actively seeking insurance and were asked the open-ended question “what prevented you from applying?” (Figure 1). In addition, patients were asked for their current occupation and the last time they had health insurance. Finally, all patients were asked to choose their “primary reason for being uninsured at this time” from a list of nine options that included an open-ended “other” category. Survey administers entered responses to open-ended questions verbatim.

Fig. 1
figure 1

Survey question sequence on experience seeking health insurance. All participants responded to both questions

Full size image

Statistical Analysis

Survey results were analyzed using SPSS software. Descriptive statistics were used for each variable. Bivariate comparisons were used to analyze responses to closed-ended questions in relation to % FPL using Chi-squared tests. For the two open-ended, free response questions: (1) “what prevented you from applying [for Medicaid]?” and (2) “what prevented you from looking [elsewhere for health insurance, including healthcare.gov]?”, two researchers coded 15 responses to each question into six categories by consensus (Fig. 4). These researchers independently sorted the remaining responses into the six categories (Fig. 5). Responses categorized as “other” included responses such as “just recently lost insurance” and “didn’t think it would be different from previous attempts.” With the responses to question one, researchers achieved a Cohen’s kappa coefficient (k) of 0.83 indicating excellent agreement [14]. With the responses to question two, researchers achieved a k of 0.63 indicating moderate agreement.

Results

Over the 3-month period that the survey was administered (March 14–June 6, 2015) UMSRFC was open for 11 clinic days, seeing 121 patients, 87 of whom provided survey responses (a response rate of 71 %). Repeat patients were removed from the data set, leaving 80 unique respondents. For each query, invalid responses were removed to give the unique “n” for that query. Of the 80 patients surveyed, 70 % of respondents with valid zip codes came from Livingston County; the next largest concentration (11 %) was from neighboring Washtenaw County (Table 1). Forty-four respondents were male and 36 were female. Ages ranged from 20 to 76 years old with a mean age of 47.7 years. Incomes ranged from 0 to 425 % FPL, with an average income just above the Medicaid cutoff of 138 % FPL. Only one respondent was above the 400 % FPL cutoff for tax credits on the HIM. Fifty-seven percent of respondents were under the income cutoff for Medicaid eligibility of 138 % FPL. Eighty percent of respondents were employed, while 15 % reported being unemployed and 5 % retired.

Table 1 Patient characteristics (n = 80)
Full size table

Only 18 % of patients had had health insurance within the last 2 years. A total of 40 % had been uninsured for over 10 years and over one-third of these had never had health insurance. A quarter of respondents reported going more than 1 year without seeing a doctor, and 8 % reported going more than 5 years.

Expense was cited most often as the primary reason for being uninsured. The next leading reasons concerned employment, including job transitions and lack of employer-provided insurance. Five out of 72 respondents (7 %) cited residency as their primary reason (Fig. 2).

Fig. 2
figure 2

Responses to “What is your primary for being uninsured at this time?” (N = 71)

Full size image

Overall, 46 out of 74 patients (62 %) surveyed had actively sought health insurance in the last 6 months through Medicaid, the Health Insurance Marketplace or elsewhere, or both. Thirty-five percent of these patients had applied for Medicaid, and 45 % elsewhere for a plan, including the Health Insurance Marketplace; 18 % of patients had explored both of these resources (Fig. 3). A total of twenty patients (25 %), had either enrolled (n = 3) or were in the process of being enrolled (n = 17), in health insurance. Medicaid application rates did not change significantly for those above and those below 138 % FPL (p = 0.901); however, individuals above 138 % FPL looked elsewhere at over twice the rate of individuals below that income level (p = 0.001).

Fig. 3
figure 3

Insurance options applied to by actively seeking individuals (N = 46)

Full size image

Twenty-eight patients had not actively sought insurance within the past 6 months. For patients who had not applied for Medicaid, the most common barriers to applying were perceived ineligibility and knowledge or resource gaps (Fig. 4). For the marketplace, knowledge and resource gaps were a barrier to many individuals as well, though the perceived cost of plans was the most frequent barrier (Fig. 5).

Fig. 4
figure 4

Barriers to applying for Medicaid among UMSRFC patients who had not applied in the last 6 months

Full size image
Fig. 5
figure 5

Barriers to looking elsewhere for insurance (including healthcare.gov) among UMSRFC patients who had not applied in the last 6 months

Full size image

Discussion

Despite eligibility under expanded Medicaid, and the availability of tax credits to nearly all respondents, over half of UMSRFC patients cited perceived expense as their primary reason for being uninsured. This finding is consistent with a 2015 national survey, where 61 % of respondents indicated that cost or expense was their primary reason for being uninsured [5]. Unlike the national survey, however, this study differentiated patients’ approaches to expanded Medicaid and the HIM. Each group, those ≤138 % FPL and those >138 % FPL, had distinct experiences and challenges seeking health insurance; these results suggest that focused outreach and education tailored to income eligibility could have a significant impact on insurance uptake among uninsured persons.

Among UMSRFC patients under 138 % FPL, almost half cited expense as their primary reason for being uninsured. Cost-sharing in Michigan’s Medicaid plans has potential to be higher than other states; however, it is designed to be no more than 5 % of an eligible individual’s income at this time [7]. Although 57 % of UMSRFC respondents were eligible for the program by income (≤138 % FPL), only 35 % had recently applied for coverage. Among the 28 respondents not actively seeking health insurance, 21 had incomes ≤138 % FPL. Further, Medicaid application rates were almost identical among persons with reported incomes above and below 138 % of the federal poverty line. This suggests a knowledge gap regarding income requirements among those apparently eligible and those apparently ineligible that may be preventing eligible individuals from accessing Medicaid. The most commonly cited reason respondents had not applied for Medicaid was that they believed they would not qualify. However, seven out of 12 who believed themselves to be ineligible were apparently eligible based on household size and income (≤138 % FPL). These findings further suggest confusion among uninsured individuals about Medicaid eligibility requirements (Table 2).

Table 2 Results of insurance seeking among UMSRFC patients actively searching in the last 6 months
Full size table

Individuals above the Medicaid eligibility cutoff were more active in seeking insurance; 68 % had recently looked elsewhere for health insurance, including healthcare.gov. However, 16 of 26 respondents in this higher-income cohort reported expense as their primary reason for being uninsured, and 17 of 22 respondents looking on the HIM had failed to find an acceptable plan. For patients above the Medicaid eligibility cutoff but below 400 % FPL, federal tax credits are designed to keep insurance premiums affordable. Unfortunately, a national survey reveals that 60 % of uninsured individuals had not heard of or were confused about the tax credits in the HIM [5]. All but one UMSRFC respondent would have qualified for such credits based on income and household size. Respondents likely considered multiple factors when assessing the total cost of insurance, including the added cost of deductibles and premiums; although tax credits do not influence deductibles, they can significantly reduce premiums for those eligible.

While other barriers to health insurance such as residency, undocumented immigrant status, and personal or political objections to the current insurance options will be difficult to overcome in the near future, this study finds expense to be a primary concern among this sample of the uninsured. Many patients reported not actively seeking insurance due to the perceived cost of HIM plans or belief they were ineligible for lower cost plans through Medicaid. Together, these results indicate that poor perceived prospects for attaining good, affordable insurance is a barrier to accessing the existing options. Education about expanded Medicaid eligibility requirements and federal tax credits may not only benefit those actively seeking insurance, but may be a way to reengage the uninsured who are disillusioned with their prospects.

This study has several limitations, including its small sample size that is partially related to the small patient load of the clinic. It is unlikely this survey offers insights on successful health insurance applicants because UMSRFC provides care to uninsured patients, and individuals who have obtained insurance are encouraged to seek care elsewhere. Additionally, using an operational definition of %FPL to estimate Medicaid eligibility inherently excludes other reasons an individual may not qualify for the program, including state and national residency requirements. All respondents who identified residency as their primary reason for being uninsured qualified as apparently eligible by %FPL, therefore potentially overestimating the proportion of Medicaid eligible patients. Additionally, the study was carried out at a free clinic where patients were receiving care, which may have altered their motivation to enroll in insurance. This effect is likely to be limited because “access to free clinics” was listed as a potential response to the primary reason for being uninsured, and no respondents selected this option.

This survey reveals important similarities and distinctions from national data that should inform further research in this area, especially in states with expanded Medicaid coverage. Some of the most prominent barriers to insurance access—perceived expense and lack of information in particular—were consistent with populations of patients at free clinics in the non-Medicaid expanding state of Utah [6]. One important of contrast is that this study took special care to emphasize Medicaid as well as HIM tax credit eligibility and access in a way that has not been addressed in national surveys or surveys conducted in non-expanding states. Further studies should assess uninsured individuals’ understandings of the cost of Medicaid and HIM plans, but also investigate what patients define as affordable. In the context of the HIM, additional studies are needed to differentiate between the financial burdens of premiums and deductibles; federal tax credits are designed to primarily mitigate premium costs, while high deductibles on certain health plans may contribute the perceived unaffordability of the HIM.

Given the importance of health insurance in determining long-term health outcomes, it is essential for those who already have close relationships with the uninsured to be guides through the changed health insurance landscape. The Robert Wood Johnson Foundation encourages outreach efforts to “continue to drill down to places where large numbers of the uninsured can be reached” [5]. If safety-net clinics such as UMSRFC can transform themselves from point of care providers into gateways to the new system of affordable insurance, they will benefit their patients as well as the health care system at large.