Article, Emergency Medicine

Return visit characteristics among patients who leave without being seen from a pediatric ED

Unlabelled imagevisit characteristics among patie”>American Journal of Emergency Medicine (2012) 30, 1019-1024

Original Contribution

Return visit characteristics among patients who Leave without being seen from a pediatric ED?

Eileen Murtagh Kurowski MD?, Terri Byczkowski PhD, Nathan Timm MD

Division of Emergency Medicine, Cincinnati Children’s Hospital Medical Center, Cincinnati, OH 45229, USA

Received 9 May 2011; revised 17 June 2011; accepted 18 June 2011

Abstract

Objectives: The primary aim of this study was to evaluate for differences in Acuity level and rate of admission on return visit between patients who leave without being seen (LWBS) and those who are initially evaluated by a physician. Our secondary aim was as well as to identify predictors of which LWBS patients will return to the ED with high acuity or require admission.

Methods: A cross-sectional study using an administrative database at an academic tertiary-care pediatric hospital in the United States from January 1, 2006, to December 31, 2008 was done.

Results: There were 3525 patients who LWBS during the study period (1.2% of total ED visits). Of these, 87% were triaged as nonurgent, and 13% as urgent at their initial visit. Two hundred eighty-nine (8%) of LWBS patients returned to the ED within 48 hours. Compared with the population who returned to the ED after previous evaluation, patients who LWBS from their initial visit and returned had significantly lower odds of urgent acuity at time of return visit (odds ratio [OR], 0.22; 95% confidence interval [CI], 0.15-0.32) and of being admitted (OR, 0.58; 95% CI, 0.40-0.84). Urgent acuity at initial visit (OR, 2.86; 95% CI, 1.35-6.04) and number of ED visits in last 6 months (OR, 1.24; 95% CI, 1.02- 1.52) were significant predictors of admission at return visit among the LWBS population.

Conclusions: Generally, patients who LWBS from a pediatric ED were unlikely to return for ED care, and those who did were unlikely to either be triaged as urgent or require hospital admission. This study showed that urgent acuity during the initial visit and number of Previous ED visits were significant predictors of admission on return. Identification of these predictors may allow a targeted intervention to ensure follow-up of patients who meet these criteria after they LWBS from the pediatric ED.

(C) 2012

Introduction

? Poster presentation: Poster presentation at the Pediatric Academic Societies Conference, Vancouver, May 2010.

* Corresponding author. Tel.: +1 513 803 6271; fax: +1 513 636 7967.

E-mail address: [email protected] (E.M. Kurowski).

The rate of LWBS visits from emergency departments (EDs) has been identified as a measure of timeliness, “one of the six dimensions of quality the Institute of Medicine established as a priority for improvement in the health care system” [1]. Previous work in general EDs describes patients who LWBS by a physician as a “high-risk” population for significant medical illness and subsequent

0735-6757/$ – see front matter (C) 2012 doi:10.1016/j.ajem.2011.06.017

deterioration [2,3]. Based on this concern, many hospitals have instituted a chart review or call-back system to reach “high-risk” patients who LWBS and ensure they have appropriate follow-up [4-6].

A study using the National Hospital Ambulatory Medical Care Survey showed a prevalence of LWBS visits of almost 2.5% for pediatric patients presenting to a pediatric ED [7]. There are, however, few studies that evaluate whether children who LWBS represent a high-risk population for deterioration after departure from the ED [6,8-10]. Also, most previous studies have focused on describing demo- graphic characteristics and reasons for departure among patients who LWBS and attempted to identify clinical deterioration by telephone follow-up [6,8-11].

The rates of LWBS visits and of unscheduled visits to the ED have both been proposed as performance measures to indicate the quality of care delivered in this setting [12]. The comparative analysis of these quality indicators provides a unique opportunity to determine if there is an overlap between patients who LWBS and those who return sicker to the ED after a prior visit. If the LWBS population is returning to the ED with higher acuity and/or requiring subsequent admission, this group could be targeted in future efforts to evaluate and improve the quality of care delivered in the ED and as a way to monitor performance improvement.

Our overall goal was to compare illness severity, as defined by urgent acuity or need for hospital admission, at time of return visit between patients who LWBS from their initial visit and those who were initially evaluated by a physician. A secondary goal was to identify predictors of patients who LWBS from their initial visit and return with high illness severity.

Methods

Study design

This was a cross-sectional retrospective study using data from a hospital administrative database. Data for 2 subsets of patients were compiled to achieve the objectives of the study. Data on all patients who LWBS were extracted to evaluate for predictors of high acuity or admission on return visit. Data on all patients who returned within 48 hours were extracted to test for differences in acuity and admission rates between patients who LWBS and those evaluated in the ED at the initial visit.

Sample size calculation

An initial sample size calculation was developed based on the population of patients who returned within 48 hours after an initial visit during which the patients either LWBS or were evaluated by a physician. All sample size calculations assumed a 2-tailed test and ? = .05 and were developed

using the software NQuery [13]. Initial sample size calculations were based on preliminary admission rate data, which showed that approximately 12% of patients who LWBS and 26% of patients who were evaluated at the initial visit returned within 48 hours and required admission. Assuming unequal group sizes, we estimated that we would need approximately 450 and 7050 in the 2 groups to have a power of 80% to detect a 5-percentage-point difference in admission rates. Next, the power associated with these initial sample sizes was estimated for testing for differences in the ordinal acuity level assuming a Wilcoxon rank sum test. Based on preliminary data, the proportions used for sample size estimates were critical/emergent (10%), urgent (25%), and nonurgent (65%). The initial sample size estimate was determined to be adequate for detecting differences of +-5% in the levels of acuity between the 2 groups. Based on annual ED volumes and that the 2 groups account for approximately 0.2% and 3.5%, respectively, of all ED visits, we estimated that this analysis would require 2.5 to 3 years of data.

Study setting and population

Cincinnati Children’s Hospital Medical Center (CCHMC) is a large, urban, tertiary-care, pediatric teaching hospital with an ED patient census of approximately 95 000 visits per year and is an American College of Surgeons-verified level I trauma center. CCHMC serves the local pediatric needs of Cincinnati, and is the primary pediatric referral center southern Ohio, northern Kentucky and southeast Indiana. The study was approved by the CCHMC Institutional Review Board.

Patient encounters from January 1, 2006, through December 1, 2008, were examined. We included all patients who LWBS and/or returned within 48 hours of initial ED visit. We excluded patients who were admitted by a route other than the ED because of the difficulty of obtaining the requisite data for this group of patients. Based on a random sample of 705 charts, we estimated the percentage admitted by a route other than the ED to be 0.16%, and thus, excluding them would not significantly impact the results of this study.

Study protocol

Data were extracted from the hospital administrative database maintained for our pediatric ED (EMSTAT Oracle 7.3.3) into a SAS, version 9.2 (SAS Institute, Inc, Cary, NC) dataset was created. Frequency distributions were developed for all variables to ensure completeness of the data and to evaluate for outliers. No outliers were identified.

Outcome measures

There were 2 primary outcome measures evaluated in our study: admission at return visit to the ED and acuity at return visit, which was used as a proxy marker for severity of

illness. Because the nonurgent acuity level accounted for more than 80% of the population included in the study, the decision was made to combine the critical/emergent and urgent categories to create a binary variable to denote 2 levels of acuity, urgent and nonurgent. Previous studies in our ED have successfully used a similar binary variable [14].

Data analysis

Frequency distributions for categorical variables and descriptive statistics for continuous variables were developed to describe the demographics and acuity of the patients who LWBS and/or returned to the ED within 48 hours.

Two logistic regression models were developed for the population of patients who returned within 48 hours to test for differences in acuity level and admission rates between those who LWBS and those who were evaluated by an ED physician on their initial visit. The dependent variables were whether or not the patient was admitted and the binary acuity level (urgent/nonurgent). The independent variable of interest was whether or not the patient LWBS. In addition, we adjusted for the following patient characteristics and measures of ED crowding by including them in the model: age, race (black, white, other), payer type (self-pay, private, government-assisted), acuity level on initial visit (urgent vs nonurgent), shift of presentation (day, evening, night), use of ED in last 6 months, ED Daily census, boarding time, and number of patients arriving to the lobby per hour.

These variables have been shown in the literature to be associated with LWBS rates [7,15-18]. Also, previous work has identifiED wait times as one of the primary motivators for patients to leave an ED without being seen and since ED overcrowding is known to prolong wait times, we adjusted for crowding variables in our model [5,16,19-20]. The three

variables we used have been evaluated previously and Race

include patients arriving to the lobby per hour, ED daily Black

60.9%

42.9%

census, and ED boarding time [21]. White

31.4%

48.8%

Other

7.8%

8.4%

Two additional logistic regression models were developed Payer

to identify predictors of acuity and hospital admission at return Commercial

17.2%

28%

visit among patients who LWBS from their initial visit. The

Government

65.6%

50%

dependent variables were admission and the binary acuity level

Self-Pay

17.2%

22%

on their return visit (urgent/nonurgent). The patient-specific

Nonurgent acuity

86.9%

60.1%

independent variables included were child’s age, race, payer

>=2 ED visits in last 6 mo

24.1%

38.3%

status, acuity level at initial visit (urgent vs. nonurgent), and number of ED visits on the last 6 months.

ED characteristics Shift of arrival

Day

27%

37.4%

Evening

60.7%

48.8%

3. Results

Night

12.4%

13.8%

on return. They did differ significantly by insurance status with more patients with a missing acuity being self-pay (42.68% vs 15.92%) and less having Public insurance (40.24% vs 66.88%). Less than 2% of the patients had missing data for any of the other predictor variables.

Table 2 shows the comparison between patients who LWBS and did not require subsequent admission and those who were admitted on return visit. The population that required admission differed significantly from those who did not in initial acuity and mean number of ED visits in the prior 6 months. Mean ED daily census and mean ED boarding time trended towards significance with P-values of .06 and

.09, respectively.

Table 3 shows that among patients who returned within 48 hours, those who LWBS were significantly less likely to be triaged with an urgent acuity level and were less likely to be admitted at return visit compared with patients who had been evaluated by a physician at the initial visit. After adjusting for both ED and patient characteristics, the odds ratio for urgent acuity and admission at return visit were 0.22 and 0.42, respectively. These models demonstrated goodness of fit using the Hosmer and Lemeshow test (P > .27). The predictive capabilities of the models were 0.66 for urgent acuity and 0.61 for admission on return as measured by the area under the Receiver operating curves.

Table 4 summarizes the predictors of hospital admission at the 48-hour return visit among the LWBS population. Admission was positively associated with urgent acuity at

Table 1 Population characteristics

All LWBS All 48-h returns (n=3525) (n = 6023)

Patient age (y), mean (range) 7.5 (0-49) 6.0 (0-27)

We identified 3525 patients who LWBS from 2006 to 2008, which represented 1.2% of the total number of visits to our ED during the study period. The characteristics of these patients are summarized in Table 1. The patients with missing initial acuity were similar to the overall LWBS population in terms of race, shift of presentation, and percentage admitted

No. of patients arriving to lobby per hour, mean (range)

ED daily census, mean (range)

ED boarding time (h), mean (range)

6.8 (0-17) 5.8 (0-18)

280 (159-408) 269.6 (116-408)

75.9 (19-331) 62.4 (16-331)

Table 2 Characteristics of LWBS patients who return and require admission

LWBS, no admit on return (n = 3490)

LWBS, admit on return (n = 35)

P

Patient age (y), mean (95% CI)

7.5 (7.3-7.7)

7.0 (4.9-9.1)

.66

Race

.11

Black

60.9

54.3

White

31.4

28.6

Other

7.7

17.1

Payer

.19

Commercial

17.3

11.4

Government

65.5

80.0

Self-pay

17.3

8.6

Initial nonurgent acuity

87.1

68.6

.001 ?

No. of ED visits in last 6 mo, mean (95% CI)

1.0 (0.9-1.0)

1.5 (1.0-1.9)

.02 ?

ED characteristics

Shift of arrival

.23

Day

27.0

20.0

Evening

60.5

74.3

Night

12.4

5.7

No. of patients arriving to lobby per hour, mean (95% CI)

6.8 (6.7-6.9)

7.6 (6.6-8.6)

.12

ED daily census, mean (95% CI)

280.3 (278.9-281.6)

293.4 (280.9-306.0)

.06

ED boarding time (h), mean (95% CI)

75.8 (74.3-77.2)

88.3 (73.5-103.2)

.90

* P b .05.

initial visit and number of ED visits in last 6 months. Age, race, and payer were not found to be significant predictors of admission on return visit in our model. This model demonstrated goodness of fit using the Hosmer and Lemeshow test (P = .56). The predictive capability of the model was 0.69, measured by the area under the receiver operating curve. There were no significant predictors of urgent acuity at the time of return visit (data not shown).

Discussion

This study provides evidence that pediatric patients who leave a pediatric ED without being seen by a physician are unlikely to be high acuity and, even when they return for care within 48 hours, are unlikely to be triaged with an urgent acuity or require admission. This study helps to provide evidence that, in a resource-limited setting, it may be advisable to prioritize a call-back or follow-up program toward the subset of patients within the LWBS population who are the most likely to return sicker. It also highlights that patients who are evaluated and subsequently return to the ED are more likely to require admission than those who return

after they LWBS at their initial visit when controlling for initial acuity.

A recent study used the National Hospital Ambulatory Medical Care Survey to identify the national prevalence of and risk factors for leaving without being seen and provides a benchmark with which we can compare our results. Specifically, they found the prevalence of LWBS for pediatric patients from a pediatric ED to be 2.46% and identified higher LWBS rates among hospitals with urban location, Self-pay insurance status, and less acute triage level [7]. Our study has an overall LWBS rate of 1.2% over the 3- year study period, which is significantly lower than the previously quoted national statistics.

Gaucher et al [15] published a study looking to identify characteristics of patients who LWBS from their pediatric ED and identified that these patients are likely to have lower triage acuity, less likely to have been referred by a physician, more likely to be age 3 months to 11 years, and reside close to the ED. This study identified a significantly higher LWBS rate (almost 17%) than ours, which raises concerns about the comparability of their results. They also did not address the outcome of patients who LWBS in their study because no follow-up information was included. It is worth noting that

Table 3 Return Visit acuity and disposition among patients who initially LWBS

Variables

Urgent acuity

Odds ratio

95% CI

Admission

Odds ratio

95% CI

Evaluated by MD LWBS

Ref

0.22 ??

0.15-0.32

Ref

0.42 ??

0.29-0.61

Adjusted for patient’s age, race, payer type, number of ED visits in past 6 months, shift of arrival, number of patients arriving to lobby per hour, ED daily census, and ED boarding time.

?? P b .001.

Variables

Admission

Odds ratio 95% CI

Patient characteristics

Age 1.00 0.95-1.05

Race

White Ref –

Black 0.98 0.43-2.20

Other 2.58 0.92-7.26

Payer

Commercial Ref –

Government 1.92 0.63-5.83

Self-pay 1.01 0.22-4.73

Urgent acuity at initial visit 3.18 ? 1.52-6.64 ?

No. of ED visits in the past 6 mo 1.24 ? 1.02-1.51 ?

* P b .05.

the acuity breakdown between Gaucher et al identified that a similar percentage of LWBS visits are low acuity when compared with our study (91% vs 87%, respectively).

A significant limitation in previous work in this area is that most studies in pediatric EDs have relied on telephone follow-up to determine the outcome of patients who LWBS. This approach is biased by not including patients without a permanent address/telephone number who are also likely to represent those without other reliable access points to the health care system. A study by Browne et al [8] from Australia attempted to overcome this by looking at admissions to their facility as well as a sample of smaller surrounding facilities for patients they were unable to contact by telephone (20% of the LWBS sample), and they were able to show that patients who LWBS were admitted at a lower rate than controls. We attempted to address this issue in a similar manner by looking at return for care to our ED among patients who LWBS. Although there are other general EDs where the patients could be taken for care, it is unlikely, given the setup of the health care system in our city, that patients with significant illness would be admitted to any other facility within the city because our institution has the only pediatric subspecialists and surgeons in the greater metropolitan area. This represents a rather unique setting in which to study rates of admission among the LWBS population without relying on telephone follow-up or primary care physician records because these are often unreliable in the medically underserved population.

In our population, we were able to identify 2 predictors of admission on 48-hour return visits among patients who LWBS: urgent triage acuity at initial visit and number of ED visits in the past 6 months. Two previous studies from general EDs have demonstrated that a larger number of previous ED visits predicted an increased likelihood to leave against medical advice or elope from their current visit [22,23]. In our patient population, approximately 24% of patients who LWBS had at

least 2 ED visits in the past 6 month, and 2.2% had at least 5 ED visits. This relationship could be confounded by underlying chronic illness, which may make patients more likely to be admitted from the ED as well as predict that they will be more frequent visitors to the ED than children without underlying conditions. Further study could be aimed at identifying the subgroup of patients who are both frequent users of the ED and more likely to be admitted to attempt to elucidate significant predictors of admission on return visit. It is worth noting that a higher percentage of the patients who were evaluated and subsequently returned to the ED (approximately 38%) had 2 or more visits in the preceding 6 months indicating that ED use might be a unifying predictor of acuity level and need for admission on return visit between the both groups. Further study would be needed to evaluate if ED use is an independent predictor of acuity and need for admission among the patients with an unscheduled return ED visit.

Our population has a different distribution of acuity levels, when compared with available US data but similar to that in Gaucher et al [15], with the breakdown of triage acuity among our LWBS population of 87% nonurgent and 13% urgent. Previous work from NHAMCS indicated that nationally pediatric patients who LWBS had the following distribution of acuity levels: 5% emergent, 31% urgent, 33% semiurgent, and 31% nonurgent [7]. Although these results are not directly comparable, because of different triage acuity scales, the NHAMCS data would appear to suggest a higher percentage of urgent visits than we identified in our LWBS population. This may be related to the lower overall rate of LWBS. Possible explanations for this include rapid triage for the more acute patients resulting in decreased wait times at our institution and subsequently decreased LWBS rates for our sicker patients. In addition, the knowledge that our facility is the only pediatric hospital in the greater metropolitan area may be an incentive for sicker patients to endure longer wait times than in other health care settings. We were unable to obtain accurate data on length of time before our patients leave before being seen by a physician so could not evaluate for this as a causative factor of our overall lower LWBS rate and lower acuity among our LWBS patients.

Table 4 Predictors of admission on return visit in patients who Initially LWBS

Because previous work has identified wait times as one of the primary motivators for patients to leave an ED without being seen and ED overcrowding is known to prolong wait times, we controlled for crowding variables in our model [5,16,19,20]. It has also been postulated that sicker patients will LWBS when crowding is worse so it was important to control for this in our model. The 3 variables we used have been evaluated previously in our ED and include patients arriving to the lobby per hour, ED daily census, and ED boarding time [21]. Emergency department patient volume has been shown previously to be associated with increased LWBS rates in general EDs [5,16,19,20]. Despite this association, we did not find a significant relationship of any of the crowding variables with acuity or need for admission at 48-hour return visit, which suggests that crowding may not cause sicker patients to LWBS.

Limitations

There are a number of limitations to the study. The first is the retrospective nature of the data collection, which limited our analysis to information that is collected routinely as part of the medical record at our institution. Ideally, we would have included information on waiting time before departure in those patients who LWBS as well as information on whether the patient has an identified primary care provider and reason for the departure [15]. Although it is possible that patients with a previous LWBS visit may be unintentionally triaged differently, if this is known at the time of acuity assignment, we feel it is unlikely in our institution for several reasons. First, to access the disposition of a previous visit, the triage nurse must open and review that visit to determine the disposition. The only information readily displayed is a flag that the patient had a visit within the past 48 hours. Second, given the volume seen at out institution, it is unlikely that the same nurse would be working in triage on both occasions and, therefore, could recognize the patient from the prior visit. Future studies could address this limitation in a prospective manner by blinding the triage nurse to the LWBS visit to determine if it alters the acuity level assigned to the patient. In addition, the results may not be generalizable to other institutions because (1) our facility is one of the few hospitals in the region that admits children and (2) we have a lower rate of LWBS visits in general compared with other pediatric EDs [7]. We also do not know how many of the children followed up with their primary care physicians and were treated accordingly, but given the nature of our pediatric Health care delivery system in our community, children requiring admission are likely to be referred to our hospital for admission, either via the ED or as a direct admission to our Inpatient unit. Important future work would be to evaluate a prospective sample of patients who LWBS from our institution to validate these predictors with the ultimate goal of a targeted call-back system for the highest-risk LWBS patients.

Conclusions

In summary, this study adds to the available literature by showing that pediatric patients who LWBS from a pediatric ED are unlikely to be high acuity and, even when they return for care within 48 hours, are unlikely to be triaged as urgent or require hospital admission. We were able to identify 2 predictors of admission on return visit, which were urgent acuity at initial visit and number of ED visits in preceding 6 months. These 2 factors could be used to identify LWBS patients who warrant intervention to help ensure adequate follow-up. In a resource-limited setting, this could save time and effort, which are currently being expended in many settings to track all patients who LWBS.

References

  1. National healthcare quality Report, 2009. AHRQ Publication No. 10-

0003, 2010. (Accessed February 22, 2011, at http://www.ahrq.gov/ qual/qrdr09.htm.)

  1. Baker DW, Stevens CD, Brook RH. Patients who leave a public hospital emergency department without being seen by a physician. Causes and consequences. JAMA 1991;266:1085-90.
  2. Fernandes CM, Daya MR, Barry S, Palmer N. Emergency department patients who leave without seeing a physician: the Toronto Hospital experience. Ann Emerg Med 1994;24:1092-6.
  3. Arendt KW, Sadosty AT, Weaver AL, Brent CR, Boie ET. The left- without-being-seen patients: what would keep them from leaving? Ann Emerg Med 2003;42:317-23.
  4. Hobbs D, Kunzman SC, Tandberg D, Sklar D. hospital factors associated with emergency center patients leaving without being seen. Am J Emerg Med 2000;18:767-72.
  5. Kronfol RN, Childers K, Caviness AC. Patients who leave our emergency department without being seen: the Texas Children’s Hospital experience. Pediatr Emerg Care 2006;22:550-4.
  6. Bourgeois FT, Shannon MW, Stack AM. “Left without being seen“: a national profile of children who leave the emergency department before evaluation. Ann Emerg Med 2008;52:599-605.
  7. Browne GJ, McCaskill ME, Giles H, Lam LT, Fasher BJ, Exley B. Paediatric walk-out patients: characteristics and outcomes. J Paediatr Child Health 2001;37:235-9.
  8. dos Santos LM, Stewart G, Rosenberg NM. Pediatric emergency department walk-outs. Pediatr Emerg Care 1994;10:76-8.
  9. Goldman RD, Macpherson A, Schuh S, Mulligan C, Pirie J. Patients who leave the pediatric emergency department without being seen: a case-control study. CMAJ 2005;172:39-43.
  10. Dershewitz RA, Paichel W. Patients who leave a pediatric emergency department without treatment. Ann Emerg Med 1986;15:717-20.
  11. Welch SJ, Asplin BR, Stone-Griffith S, Davidson SJ, Augustine J, Schuur J. Emergency Department operational metrics, Measures and Definitions: Results of the Second Performance Measures and Benchmarking Summit. Ann Emerg Med.
  12. Elashoff JD. nQuery Advisor Version 7.0. Cork, Ireland: Statistical Solutions, Inc; 2007.
  13. Timm NL, Ruddy RM. Demographics of patient visits during high daily census in a pediatric ED. Am J Emerg Med;28:56-60.
  14. Gaucher N, Bailey B, Gravel J. Who are the children leaving the emergency department without being seen by a physician? Acad Emerg Med;18:152-7.
  15. Mohsin M, Young L, Ieraci S, Bauman AE. Factors associated with walkout of patients from New South Wales hospital emergency departments, Australia. Emerg Med Australas 2005;17:434-42.
  16. Ruger JP, Richter CJ, Spitznagel EL, Lewis LM. Analysis of costs, length of stay, and utilization of emergency department services by frequent users: implications for health policy. Acad Emerg Med 2004;11:1311-7.
  17. Ding R, McCarthy ML, Li G, Kirsch TD, Jung JJ, Kelen GD. Patients who leave without being seen: their characteristics and history of emergency department use. Ann Emerg Med 2006;48:686-93.
  18. McMullan JT, Veser FH. Emergency department volume and acuity as factors in patients leaving without treatment. South Med J 2004;97:729-33.
  19. Polevoi SK, Quinn JV, Kramer NR. Factors associated with patients who leave without being seen. Acad Emerg Med 2005;12:232-6.
  20. Timm NL, Ho ML, Luria JW. Pediatric emergency department overcrowding and impact on patient flow outcomes. Acad Emerg Med 2008;15:832-7.
  21. Goodacre S, Webster A. Who waits longest in the emergency department and who leaves without being seen? Emerg Med J 2005;22:93-6.
  22. Kennedy J, Rhodes K, Walls CA, Asplin BR. Access to emergency care: restricted by long waiting times and cost and coverage concerns. Ann Emerg Med 2004;43:567-73.

Leave a Reply

Your email address will not be published. Required fields are marked *