Site of Treatment for Non-Urgent Conditions by Medicare Beneficiaries: Is there a role for Urgent Care Centers?

Abstract

Purpose

There is limited information on where and how often Medicare beneficiaries seek care for non-urgent conditions when a physician office visit is not available. Emergency departments are often an alternative site of care and urgent care centers have now also emerged to fill this need. The purpose of the study is: to characterize the site of care for Medicare beneficiaries with non-urgent conditions; the relationship between physician office, urgent care center, and emergency department utilization; and specifically the role of urgent care centers.

Methods

The study is a retrospective cross-sectional study of fee-for-service Medicare beneficiaries for FY 2012. The main outcome was rate and geographic variation of urgent care center, emergency department, or physician office utilization.

Results

Care for non-urgent conditions most commonly occurred in physician offices (65.0 per 100 beneficiaries). In contrast, urgent care centers (6.0 per 100 beneficiaries) were a more common site of care than emergency departments (1.0 per 100 beneficiaries). Overall, 83% of non-urgent visits were physician offices, 14% urgent care centers, and 3% emergency departments. There was regional variation in urgent care center, emergency department, and physician office utilization for non-urgent conditions. Areas of higher emergency department utilization correspond to areas of lower urgent care center and physician office utilization, while areas of higher urgent care center utilization had lower emergency department utilization.

Conclusions

Urgent care centers are an important site of care for Medicare beneficiaries for non-urgent conditions. There is regional variation in the use of urgent care centers, emergency departments, and physician offices with areas of low urgent care center utilization having higher emergency department utilization. The utilization of urgent care centers for treatment for non-urgent conditions may decrease emergency department utilization.

INTRODUCTION

In the U.S. health care system, gaps exist in the ability of individuals to access the primary care system in a timely and cost effective manner, particularly for non-urgent conditions, i.e. a condition that is not life-threatening or likely to result in hospitalization.¹ Often individuals either do not have a primary care physician or have trouble scheduling a timely appointment with a primary care physician.² As a result, many choose to seek care in emergency departments for non-urgent conditions not requiring emergency care.³ However, the cost of providing care for non-urgent conditions in the emergency department is significantly higher than the cost of providing care in other clinical settings.^4,5 If individuals with non-urgent conditions could be shifted to settings other than the emergency department, it is estimated $4.4 billion in costs to the health care system could be saved.³ Therefore, providing alternative care locations to emergency departments for individuals with non-urgent illnesses have both cost and quality of care implications.

Since 1980, and particularly over the last 15 years, urgent care centers have emerged to fill the need for non-urgent acute care.⁶ Broadly defined, urgent care centers provide services in episodic ambulatory care on a walk in basis, including off-hours availability.^6,7 There are now almost 9,000 urgent care centers across the country.⁶ Urgent care centers see over 70 million patient care visits per year.⁶ Urgent care centers have been shown to provide care for non-urgent conditions equivalent in quality to that provided in the emergency department at a lower cost.⁵ Urgent care center use has also been shown to decrease emergency department visits, decreasing the frequency of emergency department visits up to 48% in the 6 months following an initial urgent care center visit.⁸

Medicare beneficiaries account for at least 15% of total emergency department visits and utilize the emergency department at rates twice that of individuals with private insurance.^4,9 Medicare beneficiaries more often seek ambulatory medical care and thus could represent a population that would benefit from easier access to care and lower healthcare costs if non-urgent emergency department visits could be shifted.^5,10 However, it is not clear that emergency department visits by Medicare beneficiaries are as commonly for non-urgent conditions as for the general population.¹¹ In addition, while it has been demonstrated that Medicare beneficiaries have high emergency department utilization rates, no data are currently available regarding urgent care center utilization for Medicare beneficiaries.^2,9,12 The lack of urgent care center utilization data for Medicare beneficiaries represents a gap in our understanding of how Medicare beneficiaries meet their acute health care needs for non-urgent conditions. Therefore, it is important to understand where Medicare beneficiaries obtain care for non-urgent conditions, how they are currently utilizing urgent care centers, and the relationship of urgent care center utilization by Medicare beneficiaries on their overall use of the emergency department and physician office for non-urgent conditions. This information is important in understanding how urgent care centers may merge with other components of the healthcare system for Medicare beneficiaries. We examined the utilization of urgent care centers, emergency departments, and physician offices by Medicare beneficiaries with pre-defined non-urgent conditions.

METHODS

Study Population

This was a retrospective cross-sectional study of fee-for service Medicare beneficiaries for fiscal year 2012. The specific target population for the analysis was Medicare beneficiaries who were treated for a pre-specified non-urgent condition at an urgent care center, emergency department, or physician office during fiscal year 2012. The pre-specified non-urgent conditions were adapted from conditions identified in the literature as non-urgent conditions commonly seen in the emergency department.³ These included: upper respiratory infections; musculoskeletal conditions, strains, back pain, arthritis, contusions; urinary tract infection; and bronchitis.^3,10 The pre-specified conditions used were identified by principal diagnosis ICD-9 codes (Table 1). Provision of care at an urgent care center, emergency department, or physician office was identified using the Medicare place of service codes and CPT codes (Table 2).

Table 1.

ICD-9 codes for non-urgent conditions.

Diagnosis	ICD-9 Code
Urinary Tract infection	599.0
Upper Respiratory Infections	460–461, 463–466
Bronchitis	490–491
Contusions	920–924
Sprains	840–844, 845.1, 848
Back	846, 847
Arthritis	710–719

Table 2.

Place of service codes and CPT codes.

Place of service	Place of Service Code	CPT Code(s)
Urgent Care Center	20
Emergency Department	23
Office Visit	11
Emergency department services
New or established patient		99281–99285
Office or other outpatient services
New patient		99201–99205
Established patient		99211–99215

Data Collection and Statistical Analysis

Data was extracted from CMS data files for FY 2012; part B Carrier RIF. The primary outcome was the rate of urgent care center, emergency department, or physician office visits per 100 Medicare beneficiaries. The overall rate for the entire fee-for-service Medicare population was determined. Rates were adjusted by age, sex, and race using a regression model. Additionally, the specific rates for Medicare beneficiaries were stratified by hospital referral regions. Secondary measures used to further characterize urgent care center utilization included: ICD-9 diagnosis; Gender; Race; <70 years; Age 70–74 years; Age 75–79 years; Age 80–84 years; Age ≥85 years. The data analysis and output was generated using Base SAS software, Version 9.4 of the SAS System for Unix. Copyright © 2013 SAS Institute Inc.

RESULTS

Patient characteristics

A total of 27,971,740 fee-for-service Medicare beneficiaries were included in the analysis. Of the 27,971740 beneficiaries, there were 10,120,693 beneficiaries who were treated for at least one of the pre-specified non-urgent conditions at one of the three sites of care. Beneficiaries who sought care at more than one site during the study period were counted separately in each group. The characteristics of the beneficiaries seeking ambulatory care for non-urgent conditions at an urgent care center, emergency department, or physician office are displayed in Table 3.

Table 3.

Characteristics of Medicare beneficiaries seeking ambulatory care for non-urgent conditions at an urgent care center, emergency department, or physician office.^a

Characteristic	Urgent Care Center (n=1,426,354)	Emergency Department (n=334,841)	Physician Office (n=8,359,498)
Sex, n (%)
Male	501,523 (35)	111,817 (33)	3,038,249 (36)
Female	924,831 (65)	223,024 (67)	5,321,249 (64)
Ageb, n (%)
<70	263,387 (18)	98,352 (29)	2,093,615 (25)
70–74	266,863 (19)	86,177 (26)	2,030,943 (24)
75–79	255,356 (18)	61,762 (18)	1,640,218 (20)
80–84	255,031 (18)	45,922 (14)	1,305,927 (16)
>85	385,717 (27)	42,628 (13)	1,288,795 (15)
Race, n (%)
Black	156,928 (11)	12,140 (4)	513,282 (6)
Other	1,269,426 (89)	322,701 (96)	7,846,216 (94)

^a=Beneficiaries who were seen at more than one site of care for treatment of a non-urgent during the study period were counted separately for each site of care.

^b=age in years

Site of non-urgent care

Care for non-urgent conditions were most commonly provided in physician offices (rate 65 per 100 beneficiaries). Similarly, urgent care centers were a more common site of care for non-urgent conditions than the emergency department (rate 6.0 per 100 beneficiaries versus 1.0 per 100 beneficiaries respectively). Overall 83% of individuals received care in physician offices, 14% urgent care centers, and 3% in emergency departments.

Geographic variation

There was considerable regional variation in the rate of urgent care center, emergency department, and physician office utilization for non-urgent conditions (Figure 1a–1c). Rate of urgent care center utilization ranged from 2.6–11.5 visits per 100 beneficiaries. Rate of urgent care center utilization for non-urgent conditions was consistently higher in hospital referral regions in the eastern United States (Figure 1a). Urgent care center utilization also tended to be higher in hospital referral regions in the southwestern United States. Urgent care center utilization in the west and northwest was comparatively low as compared to other areas of the country.

Figure 1.

Utilization rates by hospital referral regions (HRR) adjusted for age, sex, and race. (a) urgent care center, (b) emergency department, (c) physician office.

Abbrev: UCC=urgent care center; ED=emergency department; OV=physician office

Rate of emergency department utilization ranged from 0.1–9.0 visits per 100 beneficiaries. In contrast to urgent care centers, rates of emergency department utilization for non-urgent conditions were higher in hospital referral regions in the western United States (Figure 1b). Although overall the distribution of emergency department utilization rates across the county appeared to be more uniform.

Rate of physician office utilization ranged from 17.1–106.0 visits per 100 beneficiaries. Rate of physician office utilization for non-urgent conditions was consistently higher in hospital referral regions in the southeastern and southwestern United States (Figure 1c). Similar to urgent care centers, physician office utilization in the west and northwest was comparatively low as compared to other areas of the country.

Areas of higher emergency department utilization tended to correspond to areas of lower urgent care center and physician office utilization. On the other hand, areas of higher urgent care center utilization tended to have lower emergency department utilization. Those areas with higher urgent care center utilization tended to have a higher ratio of urgent care center to emergency department utilization for non-urgent conditions (Figure 2a). Similarly, in areas of lower physician office utilization there was a higher ratio of urgent care center to physician office utilization for non-urgent conditions (Figure 2b).

Figure 2.

Utilization rate ratios by hospital referral regions (HRR) adjusted for age, sex, and race. (a) urgent care center to emergency department and (b) urgent care center to physician office.

Abbrev: UCC=urgent care center; ED=emergency department; OV=physician office

DISCUSSION

There are little data available on where Medicare beneficiaries obtain treatment for non-urgent conditions. This is the first study to report on the site of care for non-urgent conditions for Medicare beneficiaries as well as the geographic variation in the utilization of urgent care center, emergency department, and physician office for non-urgent conditions by Medicare beneficiaries. There are several important observations from our study regarding how Medicare beneficiaries seek care for non-urgent conditions. First, while the treatment for non-urgent conditions by Medicare beneficiaries is most common in a physician office, urgent care centers are utilized for non-urgent conditions by Medicare beneficiaries at a higher rate than they utilize emergency departments. Second, there is considerable regional variation in the utilization of urgent care centers, as well as emergency departments and physician offices. Urgent care center utilization for non-urgent conditions was inversely related to emergency department utilization as well as physician office utilization or non-urgent conditions. Taken together our finding suggest that, while physician office are the most common site of care for non-urgent conditions, urgent care centers are an important alternative site of care for Medicare beneficiaries seeking treatment for non-urgent conditions and increasing utilization of urgent care centers may in turn decrease the use of emergency departments for treatment of these conditions.

Routinely patients with acute non-urgent clinical problems seek care in emergency departments.³ It is estimated that as many as 30% of emergency department visits are for non-urgent conditions.⁴ While an individual’s perception of the acuity of their symptoms is clearly important in where they choose to seek care, lack of access to a primary care physician is the most important driver for individuals who seek alternative sites of care.^2,13 Financial considerations appear to be less important than access and perceived acuity of symptoms in an individual’s decision making.¹⁴ The goal of urgent care centers is to provide an alternative to the emergency department or no care when an individual has no primary care physician or their primary care physician is not available.^5,10 Current evidence suggests that urgent care centers provide care equivalent in quality, as measured by quality indicators, to a primary care physician or the emergency department at a lower cost to both the patient and payer and decrease emergency department visits.^5,8 More recently retail clinics have developed to fill a similar role to urgent care centers, although, in contrast to urgent care centers, they are located in retail outlets and provide more limited services than urgent care centers.¹ Similar to urgent care centers, retail clinics provide quality care for non-urgent conditions at a lower cost than emergency departments.⁵

Prior to our study little data have been available on where Medicare beneficiaries seek care for non-urgent conditions. Similarly, while Medicare beneficiaries are frequently seen in the emergency department, up until now little data has been available as to how often they use the emergency department for non-urgent conditions.^4,9 Similarly, the utilization of urgent care centers by Medicare beneficiaries is unknown. Not surprisingly we found physician offices were the main sites for non-urgent care visits. We have found that Medicare beneficiaries do utilize urgent care centers for the treatment of non-urgent conditions and the utilization of urgent care centers for non-urgent conditions is almost five-fold greater than emergency department utilization for these conditions. Our study provides an important overview of where Medicare beneficiaries obtain care for non-urgent conditions. This information is important in understanding how alternative sites can be integrated within the primary care system for Medicare beneficiaries.

It has been reported that individuals from all payers (Medicare, Medicaid, or Private insurance) all utilize the emergency department for non-urgent conditions that could be treated in an alternative setting. According to the National Hospital Ambulatory Medical Care Survey, based on the emergency department triage classification, the percent of individuals with non-urgent conditions for each payer group was: 6.8% of all Private insurance patients, 9.8% of all Medicaid patients, and 4.4% of all Medicare patients.¹⁵ There are little data on the utilization of urgent care centers by payer group. A survey by Weinick et al. from a small sample of urgent care centers that reported that the payer mix for all visits was: 50% Private Payer, 15% Medicare, and 10% Medicaid.¹⁶ In contrast, the National Hospital Ambulatory Medical Care Survey reported that the payer mix for all emergency department visits was: 35% Private Payer, 32% Medicaid, and 18% Medicare.¹⁵ While all medical groups discourage emergency department use for non-urgent conditions it is difficult to define an acceptable use rate of the emergency department for non-urgent medical conditions. This is in part due to the fact that patients with non-urgent conditions and patients with emergent conditions often have similar presenting complaints.¹⁷

The use of the emergency department for non-urgent conditions is also observed in health care systems in other countries.^18–22 The reasons for an individual choosing the emergency department for non-urgent conditions in other countries are similar to the reasons in the United States; perceived acuity of symptoms and access.^18–22 Health care systems in other countries have also looked to alternative sites to the emergency department to provide care for individuals with non-urgent conditions.^18–22

We found considerable geographic variation in the utilization rate for all sites of care: urgent care center, emergency department, and physician office. Regions with high urgent care center utilization tended to have lower emergency department utilization, while both emergency department and urgent care center utilization were higher in regions with lower physician office utilization. Although it has been shown that utilization of urgent care center will reduce emergency department use in other populations, we cannot determine whether there is a direct relationship between urgent care center utilization and emergency department utilization in our study.⁸ Medicare beneficiaries are older with more chronic illness and thus their emergency department visits for non-urgent conditions may be less preventable.¹¹ In addition, the use of discharge diagnosis to identify non-urgent conditions may not reflect how patients present to the emergency department as patients with non-urgent discharge diagnoses may look similar to higher acuity patients on presentation to the emergency department.¹⁷ Furthermore, it is possible that the geographic variability in utilization reflects availability of alternative care sites. However, the results of the study are encouraging that increasing the integration of urgent care centers into the primary care system for Medicare beneficiaries could reduce emergency department utilization in this population. Similarly, to the extent that some of the physician office visits represent visits that are delayed due to availability of an appointment, more widespread availability of urgent care centers could shift some physician office visits to a more available urgent care center.

If urgent care centers can replace the emergency department for the care of non-urgent conditions when an individual is not linked to a primary care physician or their primary care physician is not available, a significant savings to the health care system could result.³ However, concerns have been raised regarding the use of urgent care centers and retail clinics rather than a physician office. In particular, it has been suggested that use of these alternative sites could disrupt normal primary care functions.¹ This has not been well studied, however collaboration between urgent care centers and hospital systems may serve to reduce continuity of care issues.^5,7

There are several observations that deserve comment. First, there was a higher proportion of Black individuals utilizing urgent care centers as compared to emergency departments or physician offices for non-urgent care. This is in contrast to the fact that Black individuals utilize the emergency department in general at a much higher rate in comparison to other racial groups.¹⁵ A lower utilization of primary care physicians may be one reason for the higher urgent care center utilization by Black individuals.²³ Additionally, the geographic location of urgent care centers may be such that they are more accessible and therefore present an alternative to emergency departments for non-urgent care for these individuals. Other factors that are not adjusted for could also be responsible for this observation. Second, the percentage of individuals greater than 85 years of age utilizing urgent care centers for non-urgent conditions was higher than what was observed for emergency departments or physicians offices. There is no obvious explanation for this observation and the data available does not allow for any firm conjectures. It is possible that the observation may reflect the non-urgent conditions that we chose to analyze and that with a broader range of non-urgent conditions, the findings might be different. Both of these findings warrant further investigation.

There are several limitations to our study. First, we used only a limited number of non-urgent conditions; broadening the non-urgent conditions considered could alter the findings. Although, given the large number of visits captured by the discharge diagnoses chosen it is unlikely that including additional non-urgent conditions would alter the findings significantly. More importantly, non-urgent conditions were categorized based upon discharge diagnosis. It has been shown that emergency department patients with low-acuity discharge diagnoses can present with complaints similar to patients who are ultimately found to have higher-acuity conditions.¹⁷ This highlights a potential problem with the use of discharge diagnosis alone to identify individuals presenting to the emergency department with non-urgent conditions who could be shifted to alternative sites of care. Second, we did not consider the number and distribution of urgent care centers in different regions. Clearly, some of the regional variation could simply reflect availability of urgent care centers, emergency departments, and physicians, rather than individual choice. Third, we did not evaluate the effectiveness or quality of care provided at alternative sites. However, care for several non-urgent conditions have been shown to be equivalent across alternative sites of care.⁵ Fourth, we did not evaluate cost of care at alternative sites. Given the high cost of emergency department care for non-urgent conditions, evaluation of the potential for cost reduction with increase in urgent care center utilization is an important area for future research. Fifth, we did not look at utilization of retail clinics by Medicare beneficiaries. While retail clinics provide more limited services than urgent care centers, they could also provide an alternative to emergency department care for non-urgent conditions. This is another area that requires future study. Finally, we limited our study to Medicare fee-for-service beneficiaries. While this is a large population study from a national single payer, our results may not be generalizable to private payers or Medicare beneficiaries with HMOs. Further, the nature of the data available limited the analysis performed to more of an overview rather than any focus on small regional analysis.

In conclusion, we have found that while physician offices represent the most common site of care for non-urgent conditions by Medicare beneficiaries also utilize urgent care centers as a site of treatment. Medicare beneficiaries utilize urgent care centers more frequently than emergency departments as an alternative site of care to physician offices for non-urgent conditions. There appears to be significant regional variation in the site of non-urgent care for Medicare beneficiaries. In particular, in regions with higher urgent care center utilization the emergency department utilization for non-urgent condition is lower. Encouraging the use of urgent care centers for treatment of non-urgent conditions when a physician office is not an alternative may provide an effective way to reduce emergency department utilization by Medicare beneficiaries for non-urgent conditions.