Urgent Care Center and Emergency Department Utilization for Non-Emergent Health Conditions: Analysis of Managed Care Beneficiaries

Abstract

The objective was to identify predictors of utilization for urgent care centers (UCCs), emergency departments (EDs), or for both services for non-emergent health conditions among beneficiaries from a managed care organization (MCO) who resided within a 10-mile radius of UCCs. A cross-sectional design was used to analyze 2016 administrative claims data from an MCO that contracted with a UCC with 12 locations (n = 20,107). Outcome variables were number of visits to UCC, ED, or both. The MCO used the New York University ED algorithm to identify non-emergent health conditions. The Behavioral Model of Health Care Utilization was used as a conceptual framework to identify predictors in the model; age, sex, race/ethnicity, distance to UCC from residence, type of insurance, primary care physician visits, inpatient admissions, chronic conditions, morbid obesity, and smoking behavior. Generalized linear models were used to analyze the association between outcomes and predictors. About 22.7% were UCC users, 66.8% were ED users, and 10.5% used both. African Americans (incident rate ratio [IRR] = 0.95; 95% confidence interval [CI]: 0.91–0.98] were less likely to use UCCs and more likely to use the ED (IRR = 1.10; 95% CI: 1.07–1.13). Beneficiaries with multiple chronic conditions were more likely to use the ED than UCCs. Distance was not a predictor of UCC or ED usage. Utilization of UCC was low for non-emergent health conditions. African Americans and individuals with multiple chronic conditions preferred the ED to UCC for non-emergent health conditions. This study implies that MCO beneficiaries, especially the African American population, need to be informed about UCC locations and services provided.

Introduction

Non-emergent health conditions are defined as those health conditions for which a delay of several hours will not result in adverse health outcomes.¹ Urgent care centers (UCCs) have become a new model of health care facilities to treat non-emergent health conditions. These are walk-in clinics with extended office hours, on-site X-rays, and laboratory tests that treat non-urgent health conditions in an ambulatory setting outside of the traditional hospital-based or freestanding emergency department (ED).²

According to the Urgent Care Association of America, there are approximately 9000 UCCs in the United States and this number is predicted to increase in the future.³ UCCs have become part of the US health care delivery system; they prevent overcrowding in EDs,⁴ provide easy access and quick service, and charge lower cost for care than EDs.^5,6 Further, because of their extended hours of operation, UCCs have become an alternative to primary care provider (PCP) visits for some individuals. PCP visits are burdened with time constraints, long waits for appointments, and a persistent shortage of PCPs.^7
–9 With no appointments needed to seek care, UCCs address the overflow of patients from primary care as well as the ED, holding a unique position in the health care system.¹⁰

Previous literature suggests that approximately 13.7%–27.1% of all ED visits could be treated at UCCs or other alternate sites, yielding approximately $4.4 billion in savings annually.² There are mixed findings on individuals' access to UCCs and reduction in ED utilization. One study shows UCC visits decreased ED visits for non-emergent health conditions by 48% without the adverse effect of increased patient hospitalization,⁵ while another study found no association between retail clinics (eg, UCCs) and a meaningful reduction in ED utilization.¹¹ Further, in previous studies, UCC utilizers are characterized as females with fewer chronic conditions⁶ who chose UCCs based largely on convenience and more timely care than for economic reasons.¹² Managed care organizations (MCOs) have begun to collaborate with UCCs to treat non-emergent health conditions.^2,6,13 Despite the rapid growth of the urgent care industry, few studies describe specific characteristics of beneficiaries from Medicaid and Medicare MCOs who utilized UCCs.¹⁴ Most previous studies were based on national surveys or populations accessing care at a single UCC associated with a hospital, and MCOs.^{2,5
–7,10,15}

Information is lacking on simultaneous analysis of utilization of UCC, ED, or both services for non-emergent health conditions for individuals living within the fixed perimeters of UCCs. The objective of this study was to identify predictors of utilization for Medicaid and Medicare MCO beneficiaries who used UCCs for non-emergent health conditions and to compare UCC users with ED users and those using both services for non-emergent health conditions residing in the same neighborhood.

Methods

Sample

Data for this study were obtained from MCO administrative claims data for calendar year 2016. Individuals in this cross-sectional study were Medicaid and Medicare MCO beneficiaries who resided within a 10-mile radius of 12 UCC locations in southwestern Pennsylvania. These centers accepted patients from 8:00 am – 8:00 pm, 7 days a week. Mapping of the study sample and their approximate residential distance from UCCs was done using the 2015 Maptitude Mapping Software (Caliper Corporation, Newton, MA).

Health care utilization behavior for non-emergent health conditions was examined for beneficiaries identified through geo-mapping software. These beneficiaries had at least 1 UCC visit, ED visit, or visits to both services for non-emergent health conditions. The New York University (NYU) ED visits algorithm was used to identify non-emergent health conditions.¹⁶ The algorithm is valid and reliable, and positively differentiates ED visits as non-emergent and emergent.^17,18 Based on the NYU algorithm, beneficiaries with non-emergent health conditions were identified from the administrative claims database based on the International Classification of Diseases, Tenth Revision codes. Similar to the NYU algorithm, ED visits for cases involving a primary diagnosis of injury, mental health problems, alcohol, or substance abuse were excluded.^16,19 The analytic data set included beneficiaries residing within a 10-mile radius of each of the 12 UCC locations, who had at least 1 visit to a UCC, ED, or both UCC and ED for non-emergent health conditions based on the NYU ED visits algorithm (Figure 1). Information on sociodemographic, health conditions, and health care utilization from the administrative claims database also was included. The final study sample included all Medicaid and Medicare MCO beneficiaries with ages ranging from newborns to 99 years (n = 20,107).

FIG. 1.

Beneficiary inclusion criteria and classification into urgent care center users, emergency department users, or both services users for non-emergent health conditions. ED, emergency department; MCO, managed care organization; UCC, urgent care center.

Variables and measurement

Dependent variables

All MCO beneficiaries who had at least 1 visit for their non-emergent health conditions were divided into 3 groups: UCC only users, ED only users, and both UCC and ED users group. Dependent variables were the number of visits for each group.

Independent variables

Andersen's Behavioral Model of health care utilization was employed as a conceptual framework to identify factors associated with utilization of UCC, ED, or both. This model is used widely in examining health care services utilization.^20

–23 According to Andersen's model, beneficiaries' health care utilization is a function of their predisposing, enabling, and need-related factors. Predisposing factors predispose beneficiaries' health care utilization. Predisposing factors included in this analysis were age (measured as ≤21, 22–44, 45–64, and ≥65 years old), sex, and race/ethnicity (white, African American, and others). The number of beneficiaries from other racial/ethnic minority groups (ie, <0.01%-0.7% of sample) was negligible; hence, beneficiaries in these groups were aggregated and constituted the “others” racial/ethnic group. Enabling factors enable or disable beneficiaries' health care utilization. Enabling factors included in this analysis were residential distance from UCC (measured in miles), type of insurance (Medicaid/Medicare), number of PCP visits, and number of inpatient admissions during the past 12 months. Need-related factors are health conditions that require beneficiaries to utilize health care. These include number of chronic conditions (none, 1–2, 3–6, ≥7), morbid obesity (body mass index [BMI] <40/BMI ≥40), and smoking behavior (yes/no).

Previous studies have shown that health care utilization for non-emergent health conditions varies by age and sex. Females,¹⁴ younger and healthy individuals,²⁴ African Americans,¹⁴ Hispanics,¹⁴ and Medicaid beneficiaries¹⁴ are more likely to use the ED for non-emergent health conditions. In addition, individuals with shorter travel distances to the ED are more likely to have a higher number of ED visits for non-emergent health conditions.¹⁴ Also, beneficiaries who have regular visits to their PCPs and those with better health behaviors (eg, being a non-smoker, maintaining healthy weight) may be less likely to have non-emergent health conditions.

Data analysis

Descriptive characteristics were examined for beneficiaries who had only UCC visits, only ED visits, and those who accessed care at both services for non-emergent health conditions. Bivariate analyses were conducted using Kruskal Wallis tests to examine any differences between the 3 groups and predictors. Further, as the dependent variables were not normally distributed and overdispersed, 3 negative binomial regression models were conducted to examine the association between outcome and predictor variables. Results were interpreted as incident rate ratios (IRRS) with 95% confidence intervals. IRRs compare the incidence rates of the primary interest group to the comparison group.²⁴ All statistical analyses were performed using SAS version 9.4 (SAS Institute Inc., Cary, NC). Analyses were conducted by the study team based on de-identified data obtained from the MCO's analytics.

Results

Among all beneficiaries (n = 20,107) who had non-emergent health conditions, 22.7% were UCC users, 66.8% were ED users, and 10.5% used both services. Average non-emergent visits for ED only users were 1.8 (±1.5), for UCC only users were 1.3 (±1.4), and for both users were 0.52 (±0.94). Table 1 shows the descriptive characteristics for the UCC (n = 4559), ED (n = 13,433), and both (n = 2115) groups. Overall, among all those who had used services for non-emergent health conditions, average age was 28 years, and more than half were female and white. Information about distance, health insurance, and health conditions is shown in Table 1.

Table 1.

Descriptive Characteristics of Managed Care Organization Beneficiaries Who Utilized Urgent Care Centers, Emergency Departments, or Both Service Locations for Non-emergent Health Conditions

Variables and measurement	Overall (n = 20,107)	UCC users (n = 4559)	ED users (n = 13,433)	UCC and ED users (n = 2115)	P-values ^*
Predisposing Factors
Age, Mean (±SD)	27.6(±19.8)	27.1(±18.4)	27.8(±20.6)	27.5(±17.1)	0.120
Age Group
≤21 years, %	44.3	47.2	44.3	38.0	<0.001^**
22–44 years, %	33.7	33.4	31.9	45.8
45–64 years, %	17.9	16.2	19.0	14.1
≥65 years, %	4.1	3.2	4.7	2.1
Sex
Male, %	36.0	37.5	36.7	28.3	<0.001^**
Female, %	64.0	62.5	63.3	71.7
Race/Ethnicity
African American, %	43.9	30.1	60.2	41.4	<0.001^**
White, %	51.4	64.9	35.1	54.3
Others, %	4.7	5.0	4.6	4.3
Enabling Factors
Distance to UCC (in miles) Mean (±SD)	3.2(±1.4)	3.1(±1.5)	3.2(±1.4)	3.1(±1.5)	<0.001^**
<3 miles, %	43.5	46.6	42.0	43.5	<0.001^**
3–5 miles, %	49.6	46.6	50.2	50.3
>5 miles, %	6.9	6.8	7.1	6.1
Type of insurance
PA Medicaid, %	87.2	88.5	86.3	89.6	<0.001
PA Medicare, %	12.8	11.5	13.7	10.4
Number of PCP visits, Mean (±SD)	3.9(±6.4)	2.7(±3.8)	4.2(±7.1)	4.1(±5.9)	<0.001^**
PCP Visits
No, %	21.3	23.9	20.8	19.4	<0.001^**
Yes, %	78.7	76.1	79.2	80.6
Number of Inpatient Admissions, Mean (±SD)	0.20(±0.72)	0.08(±0.33)	0.25(±0.82)	0.19(±0.68)	<0.001^**
Inpatient Admissions
No, %	86.4	93.3	84.0	86.5	<0.001^**
Yes, %	13.6	6.7	16.0	13.5
Need-Related Factors
Chronic Health Conditions, Mean (±SD)	0.86(±1.16)	0.71(±1.00)	0.90(±1.2)	0.95(±1.1)	<0.001^**
0 %	50.2	59.7	48.0	44.0	<0.001^**
1–2, %	27.1	25.1	27.3	30.5
3–6, %	14.9	11.6	15.7	17.4
≥7, %	7.7	3.6	9.1	8.0
Morbid Obesity
No, %	84.9	87.3	84.2	84.1	<0.001^**
Yes, %	15.1	12.7	15.8	15.9
Smoking
No, %	92.8	93.3	92.8	91.6	0.047
Yes, %	7.2	6.7	7.2	8.4

^{^*}Results from chi-square and Kruskal Wallis Test.

^{^**}Significant at P < 0.05.

ED, emergency department; PA, Pennsylvania; PCP, primary care physician; SD, standard deviation; UCC, urgent care center.

Across the 3 groups, the majority of users belonged to the group ages ≤21 years among both UCC and ED users for non-emergent health conditions, while the group ages 22–44 years used both UCC and ED for non-emergent health conditions (Table 1). The majority of UCC users and both facility users for non-emergent health conditions were white. However, African Americans accounted for 60% of ED users for non-emergent conditions. Among UCC users, more than 40% resided both within 3 miles and 3–5 miles of a UCC and of those who used the ED for non-emergent conditions, 50% resided within 3–5 miles of a UCC. Medicaid beneficiaries were the majority users in all 3 groups. Average numbers of PCP visits were higher among those who went to the ED for non-emergent health conditions and those who used both services (4 PCP visits) than those who had UCC visits (3 PCP visits). More UCC users than ED users did not make any PCP visits for non-emergent health conditions (Table 1). More ED users than UCC users had inpatient admissions for non-emergent health conditions. About 60% among UCC users did not report any chronic conditions. Morbid obesity was higher among ED users and those who used both services compared to the UCC group (Table 1).

Table 2 shows results from negative binomial regression. UCC users were more likely to be female, Medicaid beneficiaries, and to have 1–2 chronic conditions compared to their respective reference groups. UCC users were less likely to be African American and less likely to have a higher number of inpatient admissions.

Table 2.

Results from the Generalized Linear Model Examining the Characteristics of Managed Care Organization Beneficiaries Who Utilized Urgent Care Centers, Emergency Departments, or Both Service Locations for Non-emergent Health Conditions

Variables (reference category)	UCC users (n = 4559) IRR (95% CI)	ED users (n = 13,344) IRR (95%CI)	UCC and ED users (n = 2115) IRR (95% CI)
Predisposing Factors
Age Group (ref: ≤21 years)
22–44 years	1.04 (0.98–1.08)	1.13 (1.09–1.16)^*	1.10 (1.04–1.16)^*
45–64 years	0.94 (0.87–1.00)	0.90 (0.87–0.94)^*	0.91 (0.83–1.00)^*
≥65 years	0.94 (0.81–1.09)	0.67 (0.62–0.73)^*	0.69 (0.56–0.85)^*
Sex (ref: Male)
Female	1.08 (1.05–1.12)^*	1.07 (1.04–1.10)^*	1.05 (1.00–1.11)
Race/Ethnicity (ref: White)
African American	0.95 (0.91–0.98)^*	1.10 (1.07–1.13)^*	1.02 (0.98–1.08)
Others	0.98 (0.90–1.06)	1.03 (0.96–1.09)	0.96 (0.84–1.09)
Enabling Factors
Distance to UCC (ref: >5 miles)
<3 miles	0.99 (0.91–1.08)	1.01 (0.97–1.06)	0.99 (0.87–1.11)
3–5 miles	0.94 (0.87–1.02)	1.03 (0.98–1.07)	0.99 (0.88–1.11)
Type of Insurance (ref: Medicare)
Medicaid	1.09 (1.02–1.16)^*	1.08 (1.02–1.15)^*	1.07 (0.95–1.21)
Number of PCP Visits	1.00 (0.99–1.01)	1.00 (1.00–1.01)	1.01 (1.00–1.01)
Number of Inpatient Admissions	0.92 (0.87–0.97)^*	1.03 (1.00–1.05)	0.98 (0.94–1.02)
Need-Related Factors
Chronic Health Conditions (ref: no health conditions)
1–2	1.08 (1.04–1.13)^*	1.18 (1.15–1.21)^*	1.10 (1.05–1.16)^*
3–6	1.07 (1.00–1.14)	1.45 (1.38–1.52)^*	1.32 (1.21–1.44)^*
≥7	1.07 (0.95–1.21)	1.69 (1.54–1.84)^*	1.41 (1.21–1.64)^*
Morbid Obesity (No)
Yes	1.04 (0.98–1.10)	0.97 (0.93–1.02)	1.01 (0.94–1.09)
Smoking (No)
Yes	0.97 (0.91–1.04)	1.01 (0.96–1.07)	1.00 (0.91–1.10)

^{^*}Significant at P < 0.05.

CI, confidence interval; ED, emergency department; IRR, incident rate ratio; PCP, primary care physician; UCC, urgent care center.

Beneficiaries who used the ED for non-emergent health conditions were more likely to be 21–44 years old, female, African American, and Medicaid beneficiaries, and to have 1–2 chronic conditions, 3–6 chronic health conditions, or ≥7 chronic conditions. The groups ages 45–64 years and ≥65 years were less likely to use ED for non-emergent health conditions (Table 2).

Beneficiaries who used both the UCC and ED for non-emergent health conditions were more likely to be 22–44 years old, and to have 1–2 chronic conditions, 3–6 chronic health conditions, or ≥7 chronic conditions (Table 2).

Discussion

A sample of beneficiaries from an MCO who visited the UCC, ED, or both for non-emergent health conditions were used to examine and compare beneficiary characteristics of each group. There are several key findings. First, overall low utilization of UCC was found for non-emergent health conditions. Second, African Americans used the ED instead of the UCC for non-emergent health conditions. Third, although age was not a significant factor for UCC use, ED users and those who used both services for non-emergent health conditions were younger (<45 years old). Finally, distance was not a predictor of UCC, ED, or both services use.

Although accessing care at UCCs for non-emergent health conditions is becoming increasingly popular, this study found low utilization of UCCs in the sample. One of the reasons might be differences in perception of the seriousness of health problem between beneficiaries and health care providers, which may lead to beneficiaries accessing care more at the ED than the UCC. According to Coleman et al, patients may receive care at the ED despite having access to alternative services for this reason.²⁵ Moreover, ED users and those who used both the UCC and the ED for non-emergent health conditions were sicker than beneficiaries who accessed the UCC for non-emergent health conditions. Another reason might be that the MCO contracted with the UCCs only since 2015; hence, beneficiaries may not be aware of what services UCCs provide, how to access care in these facilities, and whether the benefits are covered by their plan.

Further, this study found that African Americans were less likely to use UCCs. These findings are similar to those of a study on UCCs conducted by Denver Health Medical Center where the populations served were 15% African American compared to 50% Hispanic and 30% white.¹² Further, in the present study data, when the distance from the beneficiary's residence to the UCC was examined for African Americans, 40% resided within <3 miles and 55.3% resided within a 3–5 mile radius of the UCC; thus, the majority of African Americans reside within a 5-mile radius of the UCC. Previous studies indicate that African Americans are likely to choose ED as their usual source of care independent of health insurance.^5,26 ED use may lead to poor patient–PCP relationships, unwarranted testing and treatment, and excessive health care spending.¹ Similarly, although not significant because of their small numbers in the present analysis, the other racial/ethnic minority groups also might be adverse to using the newer UCC setting. Some of the reasons suggested for choosing the ED are long wait times for PCP appointments, lack of access to PCPs on evenings and weekends,²⁷ and convenience that includes travel time and location.¹ In addition, ED settings might be more culturally competent than UCCs at treating racial/ethnic minorities. Although UCCs may provide convenience similar to the ED, it is not clear why African Americans in this sample did not use UCC services for non-emergent health conditions. Lack of awareness or trust in UCC services, perceptions about their health problem, and quality of services may be some of the reasons. It should be noted that Medicaid beneficiaries in this sample did not incur any out-of-pocket expenses when care was received at either the ED or the UCC; however, Medicare beneficiaries were charged up to $45. Future studies should focus on examining why African Americans prefer the ED to the UCC for non-emergent health conditions when UCCs are located closer to where they reside and examining the influence of Medicaid and Medicare insurance type on association between out-of-pocket expenses and ED/UCC service use.

In addition, age was not a predictor of UCC use. However, ED users and those who used both services for non-emergent health conditions were younger (<45 years old). Older beneficiaries (≥45 years) in this sample did not access the ED for non-emergent health conditions or both locations for services. These findings on age are similar to previous studies on UCC use^7,15,28,29 and ED use for non-emergent health conditions.^1,15,28,29 Older adults prefer their PCPs than to contact off-hours services,²⁹ while the younger population preferred the ED and the UCC over a primary care practice.^15,28 According to estimates from the 2014 National Ambulatory Medical Care Survey, among adults, physician visit rates were highest for the group ages ≥65 years (589 visits/100 persons), followed by those ages 45–64 years (330 visits/100 persons) and were attributed to multiple chronic conditions.³⁰ Frequent physician visits may establish better patient–physician relationships, continuity of care, and ultimately better satisfaction. Thus, a higher disease burden and PCP visits among adults in the age ≥45 years group in this sample may be the reason for fewer ED or UCC visits for non-emergent health care. The average number of chronic conditions and PCP visits were 2 and 7, respectively, among older beneficiaries compared to 1 and 3, respectively, among those ages 22–44 years.

The present study also found that distance was not a significant factor in UCC use. The analysis included beneficiaries residing within a 10-mile radius of each of the UCC locations and average distance from the UCC was 3 miles. A reason for distance not being significant may be lack of variability in the data. About 93% resided within a 5-mile radius of the UCC. These results indicate that MCOs should contract with UCCs that have a higher volume of their beneficiaries residing within a 5-mile radius of the UCC location.

Furthermore, it is worth noting that this sample of largely Medicaid managed care beneficiaries had a low prevalence of smoking. One of the reasons might be Pennsylvania's Medicaid expansion under the Affordable Care Act that has coverage for evidence-based smoking cessation treatments for low-income beneficiaries.^31
–33 In addition, this study observed that beneficiaries with higher numbers of chronic conditions used either the ED or both services. Although higher ED use among the Medicaid and Medicare population has been documented previously,³⁴ the increased health care use among this group may indicate poor management of their health conditions. At a health system level, it indicates poor quality of care, lack of coordinated care, and inefficient use of the ED for non-emergent conditions.³⁴

These results should be considered in the light of some limitations. The beneficiaries studied were from a single urban county; hence, these results are not generalizable to other settings. However, to the best of the study team's knowledge, this is the first study to identify health care utilization behavior of MCO beneficiaries for non-emergent health conditions by mapping beneficiaries residing within a specific distance of the UCC. Future studies should focus on wider geographic areas to examine beneficiary characteristics of UCC and ED utilizers for non-emergent health conditions. The study team did not have information on the time of day when beneficiaries utilized care for their non-emergent health conditions. UCC hours of operation were from 8:00 am-8:00 pm and not 24 hours a day, so those beneficiaries who used the ED or both services may have utilized them when the UCCs in their area were closed. This analysis was conducted based on cross-sectional data; the usage may vary by year. Further, the team did not have information on whether any beneficiaries were triaged from the ED to the UCC or vice versa. This may have created substantial differences between the 2 groups and may have overestimated the results. Recent literature has shown that patients presenting to EDs have worse risk-adjusted short-term outcomes than those presenting to outpatient clinics, even after controlling for acuity level of visit or discharge diagnosis.³⁵ Existing measures of acuity using administrative data may not adequately capture severity of illness, making judgments about the appropriate setting for care difficult.³⁵ Future studies should explore the likelihood of a patient going to the ED versus the UCC for their next non-emergent health need and interventions that will be useful to sway a patient toward utilizing the UCC.

Conclusion

This study implies that although UCCs were available within a 5-mile radius for the majority of beneficiaries, including African Americans, beneficiaries may not use them for non-emergent health conditions. Access alone may not help with utilization of UCC services; interventions targeted at informing the community about UCC locations, services provided, and when they may use them, is equally important to disseminate.

Footnotes

Author Disclosure Statement

The authors declare that there are no conflicts of interest. The authors received no financial support for this article.

References

Uscher-Pines

, Pines

, Kellermann

, Gillen

, Mehrotra

. Emergency department visits for nonurgent conditions: systematic literature review. Am J Manag Care, 2013; 19:47–59.

Weinick

, Burns

, Mehrotra

. Many emergency department visits could be managed at urgent care centers and retail clinics. Health Aff (Millwood), 2010; 29:1630–1636.

Urgent Care Association of America. The case for urgent care. Published September 2011. https://c.ymcdn.com/sites/www.ucaoa.org/resource/resmgr/Files/WhitePaperTheCaseforUrgentCa.pdf Accessed December 3, 2017 .

Borkowski

. Solutions for ED overcrowding: increasing urgent care centers. J Emerg Nurs, 2012; 38:116–117.

Merritt

, Naamon

, Morris

. The influence of an urgent care center on the frequency of ED visits in an urban hospital setting. Am J Emerg Med, 2000; 18:123–125.

Thygeson

, Van Vorst

, Maciosek

, Solberg

. Use and costs of care in retail clinics versus traditional care sites. Health Aff (Millwood), 2008; 27:1283–1292.

, Hsia

. Community characteristics associated with where urgent care centers are located: a cross-sectional analysis. BMJ Open, 2016; 6:e010663.

Massachusetts Medical Society. Massachusetts patients are facing reduced access to care, longer wait times for appointments as physician shortages persist. Published 2005. http://www.massmed.org/AM/Template.cfm?Section=Research_Reports_and_Studies&TEMPLATE=/CM/Content Display.cfm&CONTENTID = 12475 Accessed December 4, 2016 .

Resnick

. Urgent care: the evolution of a revolution. Isr J Health Policy Res, 2013; 2:39.

10.

Weinick

, Bristol

, DesRoches

. Urgent care centers in the U.S.: findings from a national survey. BMC Health Serv Res, 2009; 9:79.

11.

Martsolf

, Fingar

, Coffey

, et al. Association between the opening of retail clinics and low-acuity emergency department visits. Ann Emerg Med, 2017; 69:397–403.

12.

Scott

, Batal

, Majeres

, Adams

, Dale

, Mehler

. Access and care issues in urban urgent care clinic patients. BMC Health Serv Res, 2009; 9:222.

13.

Mehrotra

, Liu

, Adams

, et al. Comparing costs and quality of care at retail clinics with that of other medical settings for 3 common illnesses. Ann Intern Med, 2009; 151:321–328.

14.

Chen

, Hibbert

, Cheng

, Bennett

. Travel distance and sociodemographic correlates of potentially avoidable emergency department visits in California, 2006–2010: an observational study. Int J Equity Health, 2015; 14:30.

15.

Coster

, Turner

, Bradbury

, Cantrell

. Why do people choose emergency and urgent care services? A rapid review utilizing a systematic literature search and narrative synthesis. Acad Emerg Med, 2017; 24:1137–1149.

16.

Billings

, Parikh

, Mijanovich

. Emergency department use in New York City: a substitute for primary care?. Issue Brief (Commonw Fund), 2000; 433:1–5.

17.

Ballard

, Price

, Fung

, et al. Validation of an algorithm for categorizing the severity of hospital emergency department visits. Med Care, 2010; 48:58–63.

18.

Gandhi

, Sabik

. Emergency department visit classification using the NYU algorithm. Am J Manag Care, 2014; 20:315–320.

19.

Billings

, Parikh

, MijanovichT. Emergency department use: the New York Story. Issue Brief (Commonw Fund), 2000; 434:1–12.

20.

Aday

, Andersen

. Models of health care utilization and behavior. In: Armitage

, Colton

, eds. Encyclopedia of biostatistics. Hoboken, NJ: John Wiley & Sons, Ltd, 2005, pp.1841–1845.

21.

Andersen

, Newman

. Societal and individual determinants of medical care utilization in the United States. Milbank Q, 2005; 83:1–28.

22.

Andersen

. Revisiting the behavioral model and access to medical care: does it matter?. J Health Soc Behav, 1995; 36:1–10.

23.

Afilalo

, Marinovich

, Afilalo

, et al. Nonurgent emergency department patient characteristics and barriers to primary care. Acad Emerg Med, 2004; 11:1302–1310.

24.

Centers for Disease Control and Prevention. Measures of association (pp 3-39–3-47). Principles of Epidemiology in Public Health Practice, 3rd ed. Published 2006. Updated May 2012. https://www.cdc.gov/ophss/csels/dsepd/ss1978/ss1978.pdf Accessed January 8, 2018 .

25.

Coleman

, Irons

, Nicholl

. Will alternative immediate care services reduce demands for non-urgent treatment at accident and emergency?. Emerg Med J, 2001; 18:482–487.

26.

Brown

, Burton

, Hixon

, et al. Factors influencing emergency department preference for access to healthcare. West J Emerg Med, 2011; 13:410–415.

27.

Durand

, Palazzolo

, Tanti-Hardouin

, Gerbeaux

, Sambuc

, Gentile

. Nonurgent patients in emergency departments: rational or irresponsible consumers? Perceptions of professionals and patients. BMC Res Notes, 2012; 5:525.

28.

Benger

, Jones

. Why are we here? A study of patient actions prior to emergency hospital admission. Emerg Med J, 2008; 25:424–427.

29.

Foster

, Dale

, Jessopp

. A qualitative study of older people's views of out-of-hours services. Br J Gen Pract, 2001; 51:719–723.

30.

National Center for Health Statistics.. Characteristics of office-based physician visits, 2014. Published December 2017. https://www.cdc.gov/nchs/data/databriefs/db292.pdf Accessed September 28, 2018 .

31.

McAfee

, Babb

, McNabb

, Fiore

. Helping smokers quit—opportunities created by the Affordable Care Act. N Engl J Med, 2015; 372:5–7.

32.

KomaJW, Donohue

, Barry

, Huskamp

, Jarlenski

. Medicaid coverage expansions and cigarette smoking cessation among low-income adults. Med Care, 2017; 55:1023–1029.

33.

American Lung Association. Stop smoking. Pennsylvania. Published 2016. http://www.lungusa2.org/cessation2/statedetail.php?stateId=42 Accessed September 28, 2018 .

34.

Peppe

, Mays

, Chang

, Becker

, DiJulio

. Characteristics of frequent emergency department users. Published 2007. https://kaiserfamilyfoundation.files.wordpress.com/2013/01/7696.pdf Accessed September 28, 2018 .

35.

Niedzwiecki

, Baicker

, Wilson

, Cutler

, Obermeyer

. Short-term outcomes for Medicare beneficiaries after low-acuity visits to emergency departments and clinics. Med Care, 2016; 54:498–503.