Patient Motivators for Emergency Department Utilization: A Pilot Cross-Sectional Survey of Uninsured Admitted Patients at a University Teaching Hospital.

, Abstract—Background: During the past several de-cades, emergency department (ED) increasing volume has proven to be a difﬁcult challenge to address. With the advent of the Affordable Care Act, there is much specula-tion on the impact that health care coverage expansion will have on ED usage across the country. It is currently unclear what the effects of Medicaid expansion and a decreased number of uninsured patients will have on ED usage. Objective: We sought to identify the motivators behind ED use in patients who were admitted to a university teaching hospital in order to project the possible impact of health care reform on ED utilization. Methods: We surveyed a convenience sample of uninsured patients who presented to the ED and were subsequently admitted to the inpatient setting. Results: Our respondents sought care in the ED primarily because they perceived their condition to be a medical emergency. Their lack of insurance and associated costs of care resulted in delays in seeking care, in reduced access, and a limited ability to manage chronic health conditions. Thus, contributing to their admission. Conclusions: Afford-ability will reduce ﬁnancial barriers to health care insurance coverage. However, efﬁcient and timely access to primary care is a stronger determinant of ED usage in our sample. Health insurance coverage does not guarantee improved health care access. Patients may continue to experience signiﬁcant challenges in managing chronic health conditions. 2015 Elsevier

, Abstract-Background: During the past several decades, emergency department (ED) increasing volume has proven to be a difficult challenge to address. With the advent of the Affordable Care Act, there is much speculation on the impact that health care coverage expansion will have on ED usage across the country. It is currently unclear what the effects of Medicaid expansion and a decreased number of uninsured patients will have on ED usage. Objective: We sought to identify the motivators behind ED use in patients who were admitted to a university teaching hospital in order to project the possible impact of health care reform on ED utilization. Methods: We surveyed a convenience sample of uninsured patients who presented to the ED and were subsequently admitted to the inpatient setting. Results: Our respondents sought care in the ED primarily because they perceived their condition to be a medical emergency. Their lack of insurance and associated costs of care resulted in delays in seeking care, in reduced access, and a limited ability to manage chronic health conditions. Thus, contributing to their admission. Conclusions: Affordability will reduce financial barriers to health care insurance coverage. However, efficient and timely access to primary care is a stronger determinant of ED usage in our sample. Health insurance coverage does not guarantee improved health care access. Patients may continue to experience significant challenges in managing chronic health conditions. Ó 2015 Elsevier Inc.

INTRODUCTION
Emergency department (ED) utilization rates have been growing rapidly in the United States, with an estimated 117 million visits in 2010 (1). Increasing ED volume is a significant challenge that prevents physicians and nurses from providing quality care to patients within a reasonable amount of time (2). Recent studies estimate that between 10% and 30% of ED visits are for nonurgent conditions (3). Such ED utilization has been attributed to barriers in timely and efficient access to primary care, lack of transportation, cost of care, wait times, and opening hours (4)(5)(6)(7)(8)(9). Some studies suggest that patients insured through the Affordable Care Act (ACA) will increase ED utilization rates (10,11).
The ACA is significantly reducing the number of uninsured patients through the expansion of Medicaid and health insurance exchanges. Medicaid eligibility and subsidized insurance may mitigate the financial barriers that contribute to increased ED use (12). However, this does not address other barriers to timely and efficient access to primary care, particularly the community-level barriers. Cheung et al. observed a direct relationship between the number of barriers faced by Medicaid recipients and ED use (8). In addition, a study by Lowe et al. indicated that community characteristics play a significant role in the decision to seek care in the ED (5). However, the attitude, behavior, and decision making of the previously uninsured will play a significant role on health care utilization patterns.
We set out, before the implementation of the ACA, to survey uninsured patients visiting the ED about their health care choices and health care utilization decisions. We sought to identify the motives for which uninsured patients sought care in the ED and project whether these would promote ED use under the ACA. We hypothesized that the ED visits in these patients were prompted primarily by emergencies, and not to fulfill primary care needs. Our ultimate goal was to describe the characteristics of a convenience sample of uninsured patients admitted to the ED at a university hospital, the impact of the lack of health insurance on their health status, and identify motivators for seeking care in the ED over of a primary care provider.

MATERIALS AND METHODS
The university teaching hospital is located in a large urban setting and provides services to inhabitants of the surrounding areas. It is the only Level I trauma center in Orange County and handles nearly 48,000 patient visits during the course of the year. We performed a descriptive, pilot cross-sectional survey administered to patients admitted to the inpatient setting through the ED between mid-July and mid-August of 2012. During this time period, medicine service case managers identified all eligible patients for this survey. Eligible patients were 18 years and older, uninsured, presented to the ED, and were subsequently admitted into the inpatient setting. These patients were then consented and interviewed by a single interviewer. The Institutional Review Board approved this study and its protocol.

Survey
We designed a 32-item survey with close-ended questions to capture the characteristics of the respondents, their insurance history, health care utilization, utilization choices, and awareness of health care resources available to them. The survey was written in English. We adopted 23 questions, with permission, from the 2007 version of the Orange County Health Needs Assessment (OCHNA) survey and the Hispanic Health Opportunity Leadership Alliance (H2OLA) mentee survey. OCHNA is a community-based not-for-profit collaborative effort that served as the primary source for health data needs from 1998 to 2012. It provided the largest health assessment of its kind at the county level in the State. H2OLA is an effort led by the California State University, Long Beach Center for Latino Community Health, Evaluation, and Leadership Training Center. Each survey required approximately 20 min to complete. For the complete survey, see the Appendix.

Study Protocol
Trained case managers identified eligible patients admitted to the inpatient setting through the ED. Verbal consent to participate in the study was obtained (see Appendix). The survey was conducted in the patients' language of choice and their responses noted by the interviewer. A single interviewer, fluent in English and Spanish, administered the survey. No identifiable patient information was obtained. Survey responses were coded and entered into a Microsoft Excel (Redmond, WA) spreadsheet, with inconsistencies referenced against the original paper survey. Our study was focused on understanding the utilization patterns of uninsured patients only and so we did not engage insured participants.

Data Analysis
For categorical measures, the distribution of subjects across each measured characteristic was described using cross-tabulation analysis. For continuous measures, descriptive statistics included the mean, median, and standard deviation, collectively providing an interpretation of the distributional characteristics of the data. STATA 13 software (StataCorp LP, College Station, TX) was used for all analysis.

RESULTS
We approached 51 patients admitted to the medicine unit of the university teaching hospital. Fifty patients agreed to participate in the study and completed the survey. One participant declined. The demographics of the respondents are shown on Table 1. Mean age of the respondents was 47 years, 60% were male, respondents were predominantly Latino (62%) and Caucasian (26%), and 16 (32%) were undocumented immigrants. They resided predominantly in North Orange County (88%). Table 2 summarizes the responses to questions about insurance history. All of our respondents expressed a desire to obtain health insurance. Fifty-eight percent of our respondents had been insured previously, but were now uninsured, primarily because of the loss of employersponsored benefits (58.6%) or loss of eligibility for government-sponsored programs (48.3%). Among those who had never been insured, 11 (52.4%) cited their legal status as the primary barrier to obtaining insurance. Despite a willingness to obtain insurance, only onethird of respondents indicated that they could afford to pay more than $20 a month as a premium, and 22% indicated that they could not afford any amount more than $5. Five (23.8%) had never tried to obtain insurance. Table 3 summarizes the responses to questions about health care utilization. About one-quarter of our respondents indicated that they had a regular source of care. Among them, the county or community clinic was identified as their source for routine care. Only one respondent identified the ED as a usual source of care. Overall, almost 60% of our respondents had sought care at a county or community clinic. However, only 62% of these were registered at the clinic. Lifestyle and dietary factors (29%) were most commonly noted as having an impact on their health, followed by lack of health insurance (25%) and stress (21%). Fifty-two percent of our

ED Utilization 205
respondents perceived that their health care utilization was less than the insured. The previously insured identified a regular source of care at a lower rate than the never insured (21% vs. 33%). Table 4 summarizes responses to questions about utilization choices. Our respondents presented at the ED primarily because they felt their condition was a medical emergency (56%) or because they were there sent by a physician (28%). Sixteen percent said they chose the ED because it accepts all patients, regardless of insurance status, and 10% described it as the fastest way to obtain care. The majority (84%) acknowledged delays in seeking care because they were uninsured, 74% felt that lack of insurance contributed to their present admission, and 86% felt that they would have been able to manage their condition better if they had insurance. The primary barrier they faced was access to primary care (54%), followed by access to medications (36%) and medical supplies (24%). Just over half (52%) of our respondents had seen a doctor within the last 6 months and almost one-third had not seen a doctor in the last year, with 58% of them citing cost as the main reason. Table 5 summarizes the results of the questions related to awareness of health care resources. Forty-two percent were somewhat or very familiar with community clinics and family health centers in their area. Among those somewhat or not familiar, 35% indicated an interest in receiving information about clinics where they could obtain low-cost health care on a regular basis. Eightyeight percent indicated an interest in receiving information about services in their area that would be beneficial to them.

DISCUSSION
Although our data might not generalizable to the entire population, our survey revealed some important themes in health care utilization, which can be used as areas of analysis in future studies attempting to address the effects of the ACA. Our study suggests that admitted uninsured patients seek care in the ED primarily because of the perceived urgency of their conditions and, less commonly, for geographic convenience, lack of health care coverage, or timely access to care. However, they acknowledged delays in seeking care and indicated that lack of insurance played a central role in not only their admission, but also their inability to manage their health appropriately. Access to primary care and the cost of medications were the main barriers arising from lack of insurance. An overall level of awareness of available community health care resources was also lacking.
About 22% of the 3 million residents of Orange County were uninsured before the expansion of Medicaid and first open enrollment of Covered California, the state health insurance exchange (13). Presently, Orange County has exceeded its enrollment target under the ACA (14). The provision of affordable health insurance options and expanded eligibility for Medicaid provides a form of health care insurance coverage for the previously uninsured. However, there are a number of dimensions of access to health care, such as availability, accessibility, adequacy, and acceptability, which are not addressed by the ACA (15). The interaction between these various dimensions determines health-seeking behavior in the population. Our survey respondents demonstrated the complex interplay between these dimensions in their health care utilization choices.
First, we would like to address that although delays in seeking care were somewhat attributed to cost by the  respondents, this was not the sole driver of utilization choices among our sample population. Thirty-four percent of the respondents cited some degree of convenience as the reason for seeking care. Among those presenting for what they felt was a nonemergency, the proximity of the location was the most commonly cited reason for their decision to seek care in the ED. The availability and accessibility of health care services are important dimensions of health care in our sample. A proportion of our patients delayed seeking care because primary care locations were geographically distant from them. An analysis of data from the American Time Use Survey shows that, on average, patients spend 35 min on travel, 42 min waiting, and 74 min receiving services. Ethnic minorities, which were a significant portion of our sample, experience longer travel and wait time than nonethnic minorities (16,17). Furthermore, four regions in Orange County, predominantly in the North, have been designated Primary Care Health Professional Shortage Areas, indicating that the primary care physician to population ratio is > 3500:1 (18). While Medicare expansion will ensure that patients have insurance coverage, it will not guarantee availability of timely health care, as the supply of physicians and physical proximity of primary care sites are not being increased proportionally. Second, access to primary care is associated with improved health status and fewer ED visits. Although increasing access is a necessary first step, the range and types of services available are just as important (19). Our data show that our respondents are largely unaware of the primary care resources in their community. More than half of our respondents indicated that they were not familiar with available community health care resources. The previously uninsured come from varying backgrounds and levels of health literacy, and are likely to have higher levels of complexity than the continuously insured. In our literature review, we identified studies of retainer programs and other health care access initiatives for the uninsured, which noted the need for patient navigators and the heightened importance of care coordination and cultural competence in this population (20).
Finally, the primary driver of ED use in our sample was the perception that their condition was an emergency. Lobachova et al. arrived at a similar conclusion in a survey of approximately 1500 ED patients (21). However, financial factors played a central role in their health status and ultimately their presentation at the ED. The majority of our respondents indicated that they delayed seeking care. Among those who had not visited a doctor in the last year, 22% cited cost as the primary reason. A large portion of our respondents also indicated that they felt that a lack of insurance contributed to their admission. Among those who indicated that a lack of insurance contributed to their ED admission, two-thirds had previously visited their local community clinic for their condition, but only 45% had visited a doctor in the last 6 months. Sixty-two percent of them cited a lack of access to primary care as the main barrier they faced due to lack of insurance, compared to just 30% among those who did not feel that lack of insurance contributed to their ED admission. Even though insurance may now be affordable, health care is not without cost. Thus, even when insured, low-income patients may still have difficulty paying for co-pays or coinsurance typically associated with private insurance.

Limitations
This study has a number of limitations. The sample size is only 50 patients and they are not representative of the general uninsured population who visit the emergency department. Compared to uninsured as measured by the 2011 to 2012 California Health Interview Survey, males and the Latino and Asian populations are overrepresented and the Caucasian population underrepresented based on the characteristics of the uninsured reported in the California Health Interview Survey 2011 to 2012 (22). However, the opinions and decisions of our respondents are strongly representative of decisions faced by the uninsured across the United States in underserved areas and large numbers of uninsured served by safety-net hospitals. Despite the small number of respondents, the themes and results are supported by the numbers presented. Although we did not collect information about the admitting diagnosis, restricting our sample to patients admitted to the medicine service excludes patients with acute  surgical conditions with greater severity. Our respondents pointed to a lack of insurance as a factor that affected their ability to manage their conditions, indicating they had chronic conditions or acute exacerbations of chronic conditions.

CONCLUSIONS
ED use among the patient sample we interviewed was driven by a perception of medical urgency, rather than by economic or convenience motives. Medical complications in this sample were attributed to delays in seeking care and lack of timely access to primary care. The provision of affordable health insurance options made available by the ACAwill increase patient health insurance coverage. However, it is unclear whether increased health care coverage will actually translate to increased access to primary care. If patients remain unable to manage their chronic conditions due to lack of appropriate access to timely primary care, then it is highly likely that ED use will increase.