Abstract
Initiatives that promote the utilization of appropriate care in appropriate settings are considered integral to achieving the Triple Aim of better care, better health, and reduced costs in the US health system.1-4 Despite progress toward achieving these aims,3,5,6 safety-net patients nevertheless face considerable barriers to accessing timely, coordinated and comprehensive health care. Safety-net patients, defined by the Institute of Medicine as low-income uninsured, Medicaid, and other vulnerable populations, 7 are more likely to receive health care that is fragmented8,9 and confront greater risks for delays in seeking/receiving care; unwarranted variations in processes/outcomes of care; being less satisfied with care; and mistrusting health care providers and systems.1,2,10-12 These disparities are associated with increased disease morbidity and mortality1,5 and reliance on emergency departments (EDs) to address unmet needs.2,5
Several strategies encourage appropriate health care utilization among safety-net populations by promoting primary care over ED use for treatment of low-acuity conditions and ongoing health needs. Initiatives that support the integration of patient-centered practices into primary care settings,13-16 such as patient-centered medical home (PCMH) implementation, have been linked to improvements in patients’ care experiences (eg, satisfaction, health care access/coordination) 17 and care processes.17-20 Other strategies to promote appropriate utilization include patient education/navigation, managed care, and financial incentives/disincentives. The effectiveness of these strategies in shifting low-acuity ED visits to primary care settings, however, is unclear.17,21-28
An underlying assumption of strategies intended to promote primary care utilization is that patients will understand and believe the implied “value proposition”—that receiving treatment in primary care settings for ongoing health care needs will deliver specific benefits over and above what they receive from EDs. However, there is evidence that this value proposition may be unclear.29-33 Several studies have found that safety-net patients’ decisions to utilize EDs for ongoing health issues represent logical, value-based choices based on distinct, multifaceted health care and social needs.29,30,33 In other words, safety-net patients may decide to use EDs because, from their perspective, doing so solves more problems and/or delivers more benefits than primary care. In this cross-sectional, exploratory study of adult safety-net patients in the southwestern United States, we examine factors motivating patients to use EDs for low-acuity conditions; describe similarities and differences by delivery setting in patients’ experiences with their usual source of care (USOC) providers; and assess patients’ awareness and perception of the PCMH concept. Our purpose is to highlight challenges and opportunities associated with advancing the value proposition of primary care for safety-net patients.
Methods
Study Setting and Participants
The institutional review boards of University of Texas Houston Health Sciences Center and (HSC-MH-11-053) and Saint Louis University (25075) approved this study, which was conducted in three safety-net hospitals in Houston/Harris County, Texas. In Houston/Harris County, Medicaid eligibility is among the most stringent in the United States, and high rates of the population remain uninsured (21.3% in Texas, 27.5% in Houston). 34 Many safety-net patients rely not only on EDs for their ongoing health care needs 35 but also receive free or discounted care from safety-net providers that include the county health system, private, not-for-profit hospitals, private physicians, community- and school-based private clinics, federally qualified health centers (FQHCs), and FQHC look-alikes (which provide similar services/programs but do not receive federal funding). 36 Like many other US cities, however, the area’s primary care capacity is severely strained and is unable to meet the full demand for primary care by safety-net populations. 36
From February to August 2012, we recruited a purposive sample of study participants from the EDs of Memorial Hermann Greater Heights Hospital, Memorial Hermann Southwest Hospital and Memorial Hermann–Texas Medical Center. Individuals eligible to participate were 18 to 64 years old, uninsured or covered by Medicaid, fluent in English or Spanish and had taken part in a patient navigation (PN) program sponsored by the Memorial Hermann Community Benefit Corporation to improve health care access for safety-net patients who were users of EDs for low-acuity conditions. 26 Patients were considered low-acuity if they were assigned an Emergency Severity Index 37 triage level of 4 or 5; when deemed appropriate by clinicians, level 3 patients were also included. As part of the PN program, community health workers educate patients about the importance of primary care, connect them with primary care medical homes and related social services, assist with appointment scheduling, and monitor/address additional barriers to ongoing use of primary care.
Data Collection
Individuals were given informed consent statements; those who verbally agreed to participate were given self-administered surveys to complete. Individuals returned the surveys in a sealed envelope and received a $25 gift card for their participation. The survey questions were adapted from relevant health care utilization and quality assessment questionnaires.38,39 Responses were measured on a 5-point Likert-type scale and categorized as strongly disagree/disagree, neither, and strongly agree/agree. We conducted cognitive interviews (n = 19) and pretesting (n = 50) in English and Spanish, implemented minor changes, and then determined the questions/format were suitable for the target population.
Measures
Patient Characteristics
We summarize the study sample using several predisposing, enabling and need factors associated with health care utilization, 40 including categorical variables (gender, age, race/ethnicity, primary language spoken, education, annual household income, insurance status) and a count of chronic conditions. We used a stepwise process to construct 3 categories of USOC status. First, we asked participants: “Is there a particular place that you usually go if you are sick and need advice about your health?” Participants who answered “no” were coded “no USOC,” participants who answered “yes” and described the person/place as an ED were coded “ED USOC”; otherwise participants were coded “non-ED USOC.”
Factors Motivating ED Use
We asked participants about several factors motivating their ED use, including urgency (“very sick,”), uncertainty (“don’t know where else to go”); paying for care (“can pay later”); convenience (“hours are more convenient); and quality (“trust the doctors,” “trust the hospital,” “more services”). We also measured participants’ agreement with the statement: “If another doctor’s office, clinic or urgent care center were available when I needed medical care, I would go to that place instead of the ER.”
Perceptions of Usual Source of Care Experiences
Participants with an ED or non-ED USOC were asked about their agreement with statements regarding access (“It is easy to contact my usual medical care provider: ‘during regular business hours,’ ‘after regular hours’”); and care coordination (“I can go to my usual medical care provider for: ‘any health problem,’ ‘preventive health care,’ ‘medication prescriptions,’ ‘referrals,’ ‘transportation . . . help applying for Medicaid’”).
Perceptions of Medical Home Concept
Drawing upon general definitions of the PCMH,13,41 we instructed participants: “One way to define a medical home is ‘a team of people led by a doctor within a medical office or clinic that serves as each patient’s primary and continuous point of contact for all health care services. The goal is to make sure patients get the care they need, are satisfied with the care they receive and have better health outcomes.’” We then asked participants about their agreement with the statement: “Every person should have a medical home,” and if they had previously heard the term “medical home.”
Statistical Analysis
Analyses were conducted using Stata version 13.1 software. 42 Descriptive statistics were calculated for patient characteristics, utilization behaviors and perceptions. Missing values were not imputed; the numbers of participants who responded to each question are reported. The non-ED, ED, and no USOC groups were compared using chi-square or Fisher’s exact tests for categorical variables and Kruskal-Wallis test for the continuous count of chronic conditions. We used a significance threshold of α = .05.
Results
A total of 329 eligible participants completed questionnaires in English (n = 260) or Spanish (n = 69) about their experiences accessing health care during the previous 12 months (Table 1). Study participants were predominantly female (67.5%); 18 to 34 years of age (52.4%); and Hispanic (51.9%) or non-Hispanic Black (37.7%). Most reported having annual household incomes of <$25 000 (85.2%). Approximately 74% were uninsured. Thirty-three percent had non-ED USOCs; 21.6% had ED USOCs; and 45.6% had no USOC.
Demographic Characteristics of Participants.
Abbreviations: USOC, usual source of care; ED, emergency department; HS, high school; GED, general education development.
n = 320 responses.
n = 311 responses.
n = 316 responses.
n = 309 responses.
There were few differences in demographic characteristics by USOC status. There were more females than males in all groups, but the proportion of females to males was highest in non-ED USOC (74.5% vs 25.5%) and no USOC groups (68.7% vs 31.3%), compared with the ED USOC group (53.7% versus 46.3%), P = .016. Mean chronic conditions were also different across groups (non-ED USOC, 1.4; ED USOC, 0.8; no USOC, 0.7; P < .001)
Table 2 reports factors motivating ED use, stratified by USOC status. Nearly all (93.4%) of participants with non-ED USOCs agreed that perceived urgency was a motivating factor, compared with 82.6% and 86.3% in ED and no USOC groups, respectively, P = .009. Sixty-two percent of those with no USOC agreed lack of knowledge about other options motivated their ED use, compared to 44.3% in non-ED and 57.4% in ED USOC groups, P = .008. Nearly 76% of participants with ED USOCs agreed that the ability to pay later motivated ED use, compared with 51.5% and 54.9% in non-ED and no USOC groups, respectively, P = .003. Convenient hours motivated 66.2% of ED and 66.9% no USOC participants, compared with 50.5% of participants in the non-ED USOC group (P = .021). Agreement was high across groups that trusting the hospital (ED USOC, 80.6%; no USOC, 76.5%; non-ED USOC, 60.8%; P = .011) and availability of more services (ED USOC, 72.6%; no USOC, 67.6%; non-ED USOC, 61.2%; P = .040) factored into their decisions to use the ED. Overall, 51.4% of participants agreed that they would utilize a non-ED place if it was available when they needed medical care; this measure was lower among the ED and no USOC groups (50.7% and 45.3%, respectively), compared with the non-ED USOC group (60.2%), P = .025.
Factors Motivating ED Use. a
Abbreviations: USOC, usual source of care; ED, emergency department; ER, emergency room.
Participants who strongly agree/agree (agree), neither agree nor disagree (neither), or strongly disagree/disagree (disagree) with the specified statement.
Table 3 reports perceptions of USOC experiences, stratified by ED or non-ED setting. The groups were similar in their agreement about ease-of-contact with their USOC during regular business hours (ED USOC, 56.1%; non-ED USOC, 57.6%; P = .664), but dissimilar in their agreement regarding ease-of-contact after-hours (ED USOC, 47.5%; non-ED USOC, 21.9%; P = .001). Compared with the ED USOC group, more participants in the non-ED USOC group agreed their USOCs would coordinate preventive care (76.0% vs 44.1%; P < .001); prescriptions (76.3% vs 45.8%; P < .001); and referrals to other medical services (67.0% vs 50.0%; P = .027). Less than half of participants (non-ED USOC, 44.1%; ED USOC, 46.4%; P = .459) agreed their USOCs would coordinate nonmedical services. Both groups expressed high agreement that every person should have a medical home (non-ED USOC, 71.0%; ED USOC, 68.6%; P = .159); 13.0% of non-ED and 22.5% of ED USOC participants (P = .224) had previously heard the term “medical home.”
Perceptions of Usual Source of Care Experiences and the Medical Home Concept. a
Abbreviations: USOC, usual source of care; ED, emergency department.
Participants who strongly agree/agree (agree), neither agree nor disagree (neither) or strongly disagree/disagree (disagree) with the specified statement.
Discussion
Strategies to increase primary care utilization assume, “if we build it, they will come,” yet increasing evidence undermines this assumption.29,30,33 In this study, only half of patients agreed they would seek treatment outside of an ED if that place was available when they needed medical care, which confirms the results of the only other study (to our knowledge) to ask a similar question of Medicaid enrollees. 29 This study also compares/contrasts perceptions regarding USOC experiences between patients with ED versus non-ED USOCs and assesses their awareness/perceptions of the PCMH concept. Although more patients with non-ED USOCs agreed that their providers helped them access needed medical services, the 2 groups reported similar experiences when asked if their providers helped them access nonmedical services related to health care. Agreement that “every person should have a medical home” was high across groups, though few were previously aware of the term.
Financial barriers (eg, strict point-of-service enforcement of co-payments in non-ED settings vs limited enforcement in EDs) and perceived quality and convenience (eg, trust ED doctors/hospitals, more services available), key factors motivating ED use in this study, are consistent with those previously reported.29,30,32,43,44 In the absence of system-level solutions, differences in payment flexibility between ED and non-ED providers are likely to persist as barriers to engaging safety-net patients in medical homes.25,45 This may be particularly relevant to patients with multiple chronic illnesses, who often have fewer personal resources but utilize more services overall (ED and non-ED), compared with those without chronic illnesses.43,46-48 Although patients with multiple co-morbidities are more likely to report having non-ED USOC providers, management of their chronic diseases may be suboptimal; as such, there are increased calls to provide more patient-centered primary care to coordinate services for high-risk populations.14-17,47 The availability of “one-stop shopping” at the ED, which may exemplify high quality for patients who associate “more” with “better” health care, and the successes of hospitals’ continuous quality improvement initiatives 30 may also confound efforts to differentiate the value proposition of primary versus ED care for safety-net patients. Additionally, some patients have reported preferring ED over non-ED USOCs due to unmet needs or negative interactions in primary care settings.10,29,32,49,50
Participants’ high agreement that every person should have a medical home suggests several opportunities to enhance the value proposition of primary care. Trusted clinicians, patient navigators and other health care team members are well-positioned to facilitate “teachable moments” in the ED about specific benefits of maintaining medical homes and to directly link patients with primary care options that meet their needs and preferences.26,51 Some organizations, including Memorial Hermann Community Benefit Corporation, have established 24-hour telephone advice lines, or implemented point-of-service interventions that navigate low-acuity patients to nearby primary care clinics in order to encourage patients to engage in continuous primary care.25,52-56 Additionally, providers can emphasize the coordination of medical and related nonmedical (eg, social) services as potentially value-added benefits of primary care, particularly given increased calls to address patients’ unmet social needs, along with barriers to health care.32,50,57
Limitations
Key limitations of this exploratory study are that we did not ask participants to rate overall access to and quality of their USOC, to rank or rate the importance of each attribute of their USOC experiences, or to name additional attributes of importance. Such measures would have enabled us to further assess differences in perceptions among those who perceived access to and quality of their USOC as good versus poor and more directly illuminated how to enhance the value proposition of primary care. These limitations should be systematically explored in future research. Other limitations include recall bias from self-reported data and the potential for selection bias inherent in nonrandom, cross-sectional studies. Furthermore, we surveyed only patients who were seeking low-acuity care in ED settings, which potentially introduces bias in the responses obtained; surveying patients in non-ED settings would have helped to address this limitation. Additionally, all participants were uninsured and Medicaid patients from one Southwestern US health system, which limits the generalizability of findings to other settings and populations.
Conclusion
Until patients perceive that seeking ongoing care in primary care versus ED settings will better serve their needs and preferences, it is unlikely that attempts to modify their utilization behaviors will be achieved and sustained over time. This study offers useful information regarding safety-net patients’ perceptions of care in ED and non-ED settings that should be further explored as opportunities to enhance and to more effectively communicate the value proposition of primary care.
Footnotes
Declaration of Conflicting Interests
The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: The lead author is a paid consultant for the Memorial Hermann Community Benefit Corporation, a not-for-profit, 501(c)(3) organization; on an as needed basis, she evaluates for the purposes of quality improvement several of its community benefit programs for safety-net populations.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the W.K. Kellogg Foundation (P0117943), the National Institute on Minority Health and Health Disparities (P60 MD000503), and the National Cancer Institute (NCI) at the National Institutes of Health (2 R25 CA57712).