Patients Who Leave Emergency Departments Without Being Seen by a Physician: Magnitude of the Problem in Los Angeles County

Patients Who Leave Emergency Departments Without Being Seen by a Physician: Magnitude of the Problem in Los Angeles County

ORIGINAL CONTRIBUTION Patients Who Leave Emergency Departments Without Being Seen by a Physician: Magnitude of the Problem in Los Angeles County From...

455KB Sizes 1 Downloads 61 Views

ORIGINAL CONTRIBUTION

Patients Who Leave Emergency Departments Without Being Seen by a Physician: Magnitude of the Problem in Los Angeles County From the Department of Emergency Medicine* and Division of General Internal Medicine, the Department of Medicine,~ UCLA School of Medicine, Harbor-UCLA Medical Center, Torrance, California; the Division of Allergy and Immunology, the Department of Pediatrics, UCLA Centerfor the Health Sciences,~ and the Department of Emergency Medicine, LAC-USC Medical Center, Los Angeles,~ California. Receivedfor publication November 16, 1992. Revisions receivedJune 7 and August 16, 1993. Acceptedfor publication September 8, 1993. Presented at the Scientific Forum of the American Collegeof Emergency Physicians in Boston, October 1991. This research was supported in part by a grant from Equality Emergency Medical Group, Los Angeles, California.

Lawrence M Stock, MD* Georgienne E Bradley~ Roger J Lewis, MD, PhD* David W Baker, MD, MPHt Jeffrey Sipsey, MD~ Carl D Stevens, MD, MPH*

Study objectives: To determine the hospital characteristics associated with patients leaving emergency departments prior to physician evaluation. Design: Cross-sectional design with data collection by mail and telephone survey. Setting: Los Angeles County, California. Type of Participants: Convenience sample of four public and 26 private hospital EDs with a combined monthly volume of 92,570.

Interventions: None. Results: Questionnaires were returned from 83% of EDs surveyed. During 1990, 4.2% of patients at these EDs left without being seen by a physician. In all, 7.3% of public hospital patients left without being seen, and 2.4% of private hospital patients left without being seen (P< .001). The percentage of patients who left without being seen was significantly higher at EDs with longer waiting times, higher fraction of uninsured patients, and at hospitals with accredited residency training programs (P< .001 for each comparison). A logistic regression model, used to simultaneously evaluate the effects of multiple correlated factors, revealed that waiting time, fraction of patients uninsured, and teaching status had independent positive associations with patients who left without being seen. Conclusion: More than 4% of patients who seek care at EDs in Los Angeles County leave without being seen by a physician. A greater proportion of patients leave without medical evaluation from EDs with long waiting times for ambulatory patients and from those that serve uninsured populations. These findings should be interpreted in light of existing data on the health consequences faced by patients who leave hospital EDs without treatment. [Stock LM, Bradley GE, Lewis RJ, Baker DW, Sipsey J, Stevens CD: Patients who leave emergency departments without being seen by a physician: Magnitude of the problem in Los Angeles County. Ann EmergMed February 1994;23:294-298.]

294

ANNALS OF EMERGENCYMEDICINE 23:2 FEBRUARY1994

LEAVE WITHOUT BEING SEEN Stoch et al

INTRODUCTION

Hospital EDs are an important entry point to the health care system in Los Angeles County, where nearly one third of more than 9 million residents lack health insurance. 1 In this setting, EDs function as a medical safety net for uninsured patients, who are often unable to gain access even to basic health services elsewhere. Two recent reports from public hospitals in California indicate that this safety net has been jeopardized by severe ED overcrowding.2,3 In the EDs studied, which generally have long waiting times, both reports found that seriously ill patients left before being seen by a physician. At one public hospital in Los Angeles County, 46% of the patients who left the ED without treatment were judged to need immediate medical evaluation, and 11% were hospitalized during the subsequent week. 3 At a public hospital in San Francisco, 4% of those who left were admitted during the 7- to 14-day follow-up period. 2 At these hospitals, long ED queuing times are associated with patients failing to gain access to medically necessary health services. Many patients who seek care at urban hospital EDs can identify no alternative site where they may see a physician) For such patients, failure to gain access to an ED may amount to a failure to receive medical treatment during a given episode of illness. Thus, data on ED access may be an important measure of the availability of ambulatory health care services for an uninsured urban population. Such information may prove useful in the current debate over health care reform. We undertook the current study to estimate the number of patients who leave EDs without treatment in Los Angeles County In particular, we wished to know whether this problem is limited to the public hospitals that care mainly for indigent patients, or whether it exists at highvolume private community hospitals as well. We also evaluated the hospital characteristics that are associated with patients leaving the ED without being seen.

estimated average ED waiting room time for an ambulatory patient arriving at 6:00 PM;percentage of patients uninsured; and hospital type (public or private; providing accredited residency training programs or not; trauma center or not). Waiting time was defined as the time from initial triage to being brought to the treatment area. In keeping with the recommendations of the American College of Emergency Physicians,4 the surveyed EDs routinely monitor the number of patients who leave without being seen as part of their ongoing quality assurance activities. Thus, survey answers were derived from existing data, except for average waiting room time, which was estimated. Surveys were completed by the medical director or a designated quality assurance coordinator. Anonymity of participating institutions was assured. The mail survey was followed by a structured telephone interview with the medical director or the quality assurance coordinator at each participating ED. The purpose of the telephone interview was to verify information on the questionnaire and to obtain information initially omitted from the returned questionnaires. Three dichotomous explanatory variables were measured: hospital type (public or private), residency program (yes/no), and trauma center (yes/no). Two additional explanatory variables, estimated average waiting time at 6:00 PMand percentage of uninsured patients, were divided into categories as follows: waiting time (less than one hour, one to three hours, or more than three hours) and percentage uninsured (0% to 25%, 26% to 50%, 51% to 75%, and 76% to 100%). Although estimates of waiting times reported by a third party are potentially unreliable, we attempted to increase the reliaFigure.

Percentage of patients who left without being seenfrorn 25 EDs No. of Hospitals 8

MATERIALS AND METHODS

A mail questionnaire was sent to medical directors of all four public and 26 high-volume private EDs in Los Angeles County. This convenience sample included all present and former adult trauma centers. The sample was chosen to allow an estimate of the overall incidence of patients leaving high-volume EDs without being seen by a physician and to allow comparisons between different types of hospitals. The questionnaire requested the following information: average monthly ED census during 1990; the average number of patients who registered for care and then left without being seen by a physician during the same period;

FEBRUARY 1994

23:2

ANNALS 0E EMERGENCY MEDICINE

7 6 5 4 3 2 1 0 0

2

3

4

5

6

7

8

9

10

Percentageof PatientsWho LeftWithout BeingSeen

295

LEAVE WITHOUT BEING SEEN

Stock et al

bility and validity of the data by using three broad categories and by having an expert (medical director or quality assurance coordinator) provide the data. The proportion of patients who left without being seen was calculated by dividing the average monthly number of patients who registered at the triage desk but left withTable 1.

1990 average monthly percentage qf patients leaving EDs without being seen by a physician, Los Angeles County, 1990 Hospital Characteristic

% Who Left Without Being Seen

Institution Type* Private

2.4 7.3

Public

Teaching* No

1.8 5.1

Yes Trauma Center* No

2.8 5.3

Yes Waiting Time (hr)* <1 1-3 >3

1.3 3.2 7.3

% Uninsured* 0-25 26-50 51-75 76-100

2.1 2.8 7.3

*P< .001 byz 2.

out seeing a physician, by the average monthly total of patients who registered for care in each ED. The Z2 test was used to test categorical variables for univariate differences in the proportion of patients who left without being seen. A multivariate logistic regression model was used to determine the simultaneous separate contribution of each potential explanatory variable to the fraction of patients who left without being seen. Statistical analysis was performed using SAS data analysis software. 5,6 RESULTS

The survey was returned by 28 of 30 (93%) EDs, and 25 of these 28 (89%) provided complete data. All four public hospitals and 21 of 26 (81%) private EDs surveyed were included in the final analysis. Ten of the 11 (91%) current adult trauma centers in Los Angeles County participated. Based on a recent estimate of the total average monthly ED visits in Los Angeles County, 7 the study sample included approximately 44% of all ED visits that occurred in the county during the study period. During calendar year 1990, an average of 92,570 patients registered to be seen each month in the participating EDs; 4,091 (4.2%) of these patients left without being seen by a physician. At individual hospitals, the proportion leaving without treatment varied from 0.01% to 9.4% (Figure). On average, 7.3% of registrants left public hospital EDs without being seen. This was more than three times the average of 2.4% reported by private hospitals (Table 1).

Table 2.

Logistic regression model for percentage of patients leaving EDs without treatment in Los Angeles County, 1990 Explanatory Variable Institution type Private

Coefficient

SE

Standard Coefficient

-1.234"

0.172

-0.328

Odds Ratio (95% Confidence Interval) 1.00 (Definition) 0.29 (0.21 - 0.41)

Public

Teaching status

0.588*

0.064

0.146

Nonteaching Teaching

Estimated wait (hr)

1.00 (Definition) 1.80 (1.58 - 2.05)

0.941"

0.065

0.411

<1 I -3 >3

1.00 (Definition) 2.56 (2.25 - 2.92) 6.56 (5.76 - 8.52)

Fraction uninsured 0.00 0.26 0.51 0.76

-

0.370*

0.057

0.265

0.25 0.50 0.75 1.00

Trauma status

1.00 (Definition) 1.45 (1.29 - 1.62) 2.10 (1.67 - 2.63) 3.04 (2.16 - 4.27)

0.014t

0.046

0.004

Nontrauma center Trauma center

1.00 (Definition) 1.01 (0.93- 1.11)

• P< .001. * Not significant (P> .05),

296

ANNALS OF EMERGENCY MEDICINE

23:2

FEBRUARY 1994

LEAVE WITHOUT BEING SEEN Stock et al

By univariate analysis (Table 1), the proportion of patients who left without being seen was significantly higher at teaching hospitals (5.1%) than at nonteaching hospitals (1.8%), and at trauma centers (5.3%) than at nontrauma centers (2.8%) (P < .001 by %2 for both). As the estimated waiting room time went from less than one hour, to one to three hours, to more than three hours, the percentage of patients who left without being seen went from 1.3% to 3.2% to 7.3% (P < .001 by Z2). As the percentage of uninsured patients went from 0% to 25%, to 26% to 50% to 76% to 100%, the percentage leaving without being seen went from 2.1% to 2.8% to 7.3% (P < .001 by Z2). The univariate analysis indicated significantly higher left-without-being-seen rates at institutions with the following characteristics: public hospitals, teaching hospitals, trauma centers, hospitals with long estimated ED waiting room times, and hospitals serving a high percentage of patients without insurance (Table 1). Determining the separate contributions of these characteristics to the number of patients who leave without being seen is difficult because the institutions with the longest estimated waiting times were the four public hospitals, all of which support accredited residency training programs and serve populations with the highest percentage of uninsured patients. In order to address this problem of covariance, a multivariate logistic regression model, incorporating all explanatory variables, was used to determine which variables were independently associated with increases in the percentage of patients who left without being seen (Table 2). The odds ratio is the probability of a patient leaving without being seen, divided by the probability of a patient waiting for treatment. In the logistic regression model, the odds ratio associated with a particular hospital characteristic is multiplied by the baseline odds ratio to obtain the odds ratio for leaving without treatment for a hospital that has that characteristic. The baseline odds ratio applies to a hospital with none of these characteristics, in other words, a private, nonteaching, nontrauma-center institution with an average wait of less than one hour and less than 25% of uninsured patients. Using this model, institution type (public versus private), teaching status, estimated waiting time, and fraction of uninsured patients were found to have statistically significant independent associations with the proportion of patients who left. Trauma center status did not have a significant independent effect when the effects of other variables were taken into account. Interestingly, public hospital status was associated with a decreased likelihood of patients leaving without treatment, after accounting for

FEBRUARY 1994 2 3 : 2 ANNALS OF EMERGENCY MEDICINE

other factors such as waiting time. Of the variables studied, estimated waiting time had the strongest independent association with the percentage of patients leaving without treatment, as demonstrated by the relative magnitudes of the standardized coefficients in the model. DISCUSSION

The results of this study demonstrate that the problem of patients leaving EDs without treatment is not limited to the two public hospitals that have recently reported this phenomenon. 2,3 It is a more general problem that affects the majority of the EDs studied. Although the problem is far more severe at public hospitals, where more than 7% of all registrants left without being seen, significant numbers of patients leave from high-volume private hospital EDs without medical evaluation as well. It should be noted that our sample did not include less busy EDs. Thus, our results cannot be extrapolated to low-volume facilities. Assuming that long waiting room times are the result of ED overcrowding, the strong association between waiting room time and the percentage of patients who left without treatment suggests that overcrowding plays an important role in patients' failure to gain access to care in EDs. The finding of average estimated waiting room times of more than three hours at the public hospitals confirms earlier reports of severe overcrowding at these facilities. 2,3,8 In addition, our results indicate that the capacity of some private hospital EDs to provide timely care for all who seek it may have been exceeded. More than 3% of registrants left without medical evaluation from those private community hospitals studied with estimated ED waiting room times of one to three hours. It is interesting to note that the results of the logistic regression model suggest that the rate of patients leaving without being seen at public facilities is lower than that at private facilities, after accounting for the effects of the other variables studied (eg, waiting time, percentage of uninsured patients). This may reflect the expectation of longer waiting times on the part of patients seeking care at public facilities. The significance of patients leaving EDs without medical treatment lies in the medical consequences that result when they fail to gain access to care in the ED. If these patients are seriously ill and unaware of options for obtaining care at sites other than the ED, then the large number of patients who leave EDs in Los Angeles County without being seen indicates that many people cannot obtain access to necessary health services. If, on the other hand, the patients who leave without being seen have

297

LEAVE WITHOUT BEING SEEN Stock et al

minor complaints and can identify other options for obtaining care, then long waiting times at EDs may present more of an inconvenience than a danger to patients who fail to gain access. This study collected no data on the illness severity of patients who left without treatment or on the alternative sources of care that were available to them. However, two recent investigations found that many patients who left public hospital EDs without treatment were severely ill and in need of immediate care. 2,3 In the study by Baker and others, 3 two thirds of the patients who left without being seen could identify no alternative site of care that would be available to them if the public hospital ED were to close. The results of these studies conducted at public hospitals may not be generalizable to the hospital sector as a whole. We found no published studies that provide direct information on illness severity or alternatives for gaining access to care among private hospital ED patients. Therefore, the medical consequences faced by patients who leave private hospital EDs are unknown. There may be important differences between public and private hospital EDs in terms of case mix, triage practices, patient expectations of waiting times, and reasons for leaving without treatment, and these factors may profoundly affect the medical outcomes of patients who leave from these different settings. Even without such data, however, in a county where one person in three lacks health insurance, it seems likely that at least some of the nearly 1,400 patients who left the surveyed private hospital EDs without being seen each month were both seriously ill and had few options for receiving care elsewhere. One important difference between the public and private setting is the role that some managed health care organizations play in controlling access to ED care. Medical directors of private EDs reported that a common reason for patients leaving their facilities without being seen is failure of managed care plans to preauthorize payment for the visit. These authorization decisions are often based on information obtained over the telephone by a physician or nurse who has not had direct contact with the patient. The accuracy of these telephone triage decisions is unknown, but clearly they could adversely affect medical outcomes if they fail to identify patients in need of immediate care. The hospital characteristics that this study found to be associated with an increased percentage of ED patients leaving without treatment are common at urban teaching hospitals throughout the United States. 8-13 It seems likely that many of these facilities, which have long waiting times and a high proportion of uninsured patients, also

298

have substantial numbers of ED registrants leaving without being seen by a physician. CONCLUSION

Large numbers of patients leave public and busy private EDs prior to medical evaluation in Los Angeles County. Long waiting room times had the strongest association with a higher percentage of patients leaving without being seen. In combination with previous studies on illness acuity and health outcomes of patients who left public hospital EDs without being seen by a physician, this study suggests that long ED queuing times are associated with patient failure to obtain needed health care. REFERENCES 1. Brown ER, Valdez RB, Morgenstern H, et ah Health InsuranceCoverageof Californians in 1989. California Policy Seminar, University of California, 1991, p 1-4. 2. Bindman AB, Grumbach K, Keane D, et al: Consequencesof queuing for care at a public hospital emergencydepartment. JAMA 1991;266:1091-1096. 3. Baker DW, Stevens CD, Brook RH: Patients who leave a public hospital emergencydepartment without being seen by a physician: Causesand consequences.JAMA 1991;266:1085-1090. 4. American College of EmergencyPhysicians: QuafityAssuranceManual for Emergency Medicine. Dallas, ACEP, 1988. 5. SAS Institute Inc: SAS/STATUser'sGuide,Release6.03Edition.Cap/, North Carolina: SAS Institute Inc, 1988. 6. SAS Institute Inc: SUGISupplementalLibraryUser'sGuide, Version5 Edition.Cap/, North Carolina: SAS Institute Inc, 1986. 7. California Department of Health Services, Data users support group, Office of Statewide Health Planningand Development,1991. 8. Andrulis DP, Kellermann AL, Hintz EA, et el: Emergencydepartments and crowding in United States teaching hospitals. Ann EmergMed 1991;20:980-986. 9. Lynn SG, KellermannAL: Critical decision making: Managing the emergencydepartment in an overcrowded hospital. Ann EmergMed 1991;20:287-2£2. 10. American College of EmergencyPhysicians:Hospital and emergencydepartment overcrowding. Ann EmergMed 1990;19:336. 11. PaneGA, FarnerMC, Salness KA: Health care access problems of medically indigent emergencydepartment walk-in patients. Ann EmergMed 1991;20:730-733. 12. Lowe RA, Young GP, Reinke B, et ah Indigent health care in emergency medicine: An academic perspective. Ann EmergMed 1991;20:790-794. 13. Dedet RW, Nishio DA: Refusing care to patients who present to an emergencydepartment. Ann EmergMed 1990;19:262-267.

Reprint no. 47/1/52672 Address for reprints: Carl D Stevens,MD, MPH ValueHealthSciences 2400 Broadway,Suite 100 SantaMonica,CA90404

ANNALS OF EMERGENCY MEDICINE

23:2

FEBRUARY 1994