~ ~

Social Work, Wrenn, Rice 247

Social-work Services in an Emergency Department: An Integral Part of the Health Care Safety Net

Keith Wreniz, MD, Nancy Rice, CSW MSW

ABSTRACT

Objective: To quantify and describe the activities of social workers dedicated to a large urban emergency department (ED).

Methods: A retrospective case series of all patients seen by social workers in an urban university hospital ED over a period of six weeks.

Results: Social-work service was provided to 5% of ED patients. Three distinct groups of patients were usually seen by social workers: the elderly, young adults, and children less than 5 years of age. The median age of the group referred for social work services was greater than that of the ED population as a whole; triage acuity also was greater in the referred group. The types of services provided varied with age. Among those patients with social-work consultations, the average time spent with each patient was over one hour and did not vary according to the age, sex, race, or insurance status of the patient. More than 60% of the social worker's time was spent with patients or their significant others.

Conclusion: Social workers provide valuable services to ED patients. The availability of social workers in the ED reduces the demands for emergen- cy physicians and nurses to arrange home health care, nursing home placement, and other social-service functions. Cost savings through diversion of nonacute social admissions are possible. The types of services provided vary and depend to a large extent on patient age. The availability of dedicated social-work personnel in the ED and the educa- tion of emergency personnel regarding the services that they can provide should be beneficial for patients, staff, and the hospital served. Acad. Etnerg. Med. 1994; 1~247-253.

248 ACADEMIC EMERGENCY MEDICINE MAY/JUNE 1994 VOL 1/NO 3

I Much has been written about societal and medical care problems besetting emergency departments (EDs), such as hospital overcrowding, child abuse and neglect, and aban- donment of the elderly, all problems related to the fact that EDs are “the ultimate safety net.”’-’ Long-term solutions to these problems must be sought on societal levels. In the short term, however, many things can be done at the level of the individual ED to cope with these problems, not the least of which is provision of comprehensive social-work ser- vices. Although there have been several descriptions of ED social-work programs in the social-work literature,*-17 almost none of these includes data about the number of patients seen, the characteristics of these patients, the types of patient problems seen by social-work staff, and the services provided. This study quantifies and describes social-work services in an ED in a medium-sized north- eastern city.

I RESULTS

Study Population Demographics

Social-work service was provided to 5% of the ED patients (560 patients) during the six-week period. During the same period, a total of 10,740 patients were seen in the ED. The population served by the social workers has an unusual trimodal distribution (Fig. l), with peaks in the ranges of the very young (< 5 years of age), young adults (25-40 years of age), and the very old (2 70 years of age). There were obvious differences in the problems identified within those age ranges. Among those older than 70 years of age, specific problems included the need for home care services, adult protective services for possible elder abuse, financial or resource management, bereavement services, or help with compliance with medical regimens. Among those between 25 and 40 years of age alcohol/drug abuse

I I 1 : O 50 100

Age (years)

FIGURE 1. Ages of the 560 patients referred for social-work consul- tation over a six-week period in the ED.

referral, crisis intervention, and domestic violence were the most common problems. Among children younger than 5 years of age, the most common problem by far was related to suspected child abuse and neglect.

Fifty-eight percent (318) of social-work referrals in the ED were for women; this proportion is slightly higher than the overall proportion of women (50%) seen in the ED over the study period (p < 0.05). The mean age of the referred women (58 years) was greater than that of the referred men (53 years) (p < 0.05, 95% CI for difference between means 0.01-10.2). There was a significant preponderance of women only in the referred group older than 85 years of age (p < 0.051, which was in proportion to their frequency among ED patients in this age group. Similarly, there was a significant preponderance of whites in the referred group older than 79 years of age (p < 0.05), which also was in proportion to their frequency among ED patients in this age group.

The racial breakdown of patients referred to the social worker was 336 (65%) white, 51 (10%) Hispanic, and 129 (25%) black. When compared with the ED population as a whole, there was an overrepresentation of white and His- panic patients and an underrepresentation of black patients.

There were more patients with higher grades of illness acuity (level 3 and above) in the population receiving social-work interventions than among the ED population as a whole (Table 1).

Time Total times spent working on individual patients’ prob-

lems averaged 64 minutes, with a median time of 50 minutes (IQ 30-85 minutes). This time was related to a variety of service interventions (Table 2). The most time- consuming task involved psychological counseling and postdischarge planning (median time of 30 minutes). The most common task was chart review and documentation, followed by information/referral, risk evaluation, and psy- chosocial assessment. The total times spent did not vary according to age, sex, race, or insurance status. With the exception of interpreter services, the times spent perform- ing specific tasks did not vary according to age, sex, race, or insurance status.

The social workers saw approximately twice as many patients during the day shift (mean of 11 2 4 patients) as during the evening shift (mean of 5 ? 2 patients). On average, 64 patients were seen in the ED each day shift during the study and 74 each evening shift. During the standard work week from Monday through Friday, approx- imately equal numbers of patients were seen each day (means from 14 to 16 k 1 to 2 patientdday). On weekends the load was about half that during the regular work week, averaging 6 to 7 & 3 patientdday (p < 0.05 for differ- ences between means).

Social Work, Wrenn, Rice 249

I METHODS

Study Design We performed a descriptive analysis of all patients

who were seen by a social worker at the University of Rochester, University Hospital ED over a six-week period (October 8 through November 19, 1992).

Facility and Patient Population The ED was staffed by emergency medicine faculty

(two-thirds of whom were diplomates of the American Board of Emergency Medicine) and residents from the Departments of Medicine, Surgery, Obstetrics and Gy- necology, and Pediatrics. The population of Rochester, New York, is approximately one million. The study population included all patients referred for social-work interventions in the ED of an urban university hospital with approximately 60,000 visits per year.

Patients who received social-work services were identified in several ways: ED staff referral of patients to social work; social-work staff review of the ED registration log for high-risk indicators (i.e., patient age of 70 years or older; patient age of 5 years or younger with injury or ingestion; or chief complaint indicating substance abuse, violence, or postrelease functional impairment); patient self-referral; and social-work staff observation of patient or family member actions indica- tive of psychosocial problems. Social work consultation from ED staff occurred either sporadically at the discre- tion of a physician or nurse or during daily morning rounds in the ED, which were attended regularly by the social workers.

Social Workers The social-work staffing in the ED during the study

period consisted of 3.5 full-time equivalent positions (FTEs) primarily assigned to the ED. Two social work- ers covered the service from 7:45 AM to 6:OO PM Monday through Friday. One social worker covered the service from 5:OO PM to 12:OO AM Monday through Friday and eight hours on seven holidays each year. One-half social worker covered the service from 9:30 AM to 8:OO PM on

Saturday and Sunday in a 0.5-FTE position. Most of the social workers had master’s degrees.

Data Collection Many of the data were collected prospectively by the

social workers themelves (e.g., times spent with pa- tients in various service activities, the global social service effort rendered, and where and when the pa- tients were seen). When possible, these data were col- lected for every patient referred. Demographic data such as age, sex, race, type of insurance, and acuity of illness were collected retrospectively from our comput- erized ED log. Also collected from the ED log were denominator data for the ED population as a whole during the study period. Illness acuity on admission to the ED was routinely graded for all patients by the triage nurses according to a five-point scale as follows: 1 (not sick), 2 (ambulatory), 3 (ill but relatively stable), 4 (unstable), and 5 (cardiopulmonary arrest). Actual nurs- ing time spent with a patient was a significant determi- nant of patient illness acuity.

Financial data were obtained retrospectively from the hospital billing office for all patients who were admitted to the hospital for nonacute social problems during the year in which the study was done. Further billing information comes from an analysis of all hospi- tal admissions.

Data Analysis Descriptive statistics are reported. Normally dis-

tributed variables are reported as mean .t standard deviation; skewed data are reported as median with interquartile range (IQ 25-75%). Chi-square analysis was used to compare differences between proportions, the Student’s t-test was used to compare differences between means of normally distributed variables, and the Wilcoxon signed-rank test was used to compare differences between medians and other nonparametric variables. The 95% confidence intervals were deter- mined for selected statistics. A p value of < 0.05 was considered significant. When individual data were missing, the denominator of the reported proportion was adjusted to represent the number of patients with available data rather than the total number of patients.

~~~ ~ ~~ ~~~ ~ ~~ ~

250 ACADEMIC EMERGENCY MEDICINE MAY/JUNE 1994 VOL 1 /NO 3

TABLE 1 Services in the ED during a Six-week Period

Demographic Data for 560 Patients Receiving Social

ED Population

( n = 560) ( n = 10,740) Study Population as a Whole

NO (Yo) 195% CI) No (70) 195% CI]

Sex Male 231 (42) [38-46]* 5,359 (50) [49-511* Female 318 (58) [54-621' 5,380 (50) [49-51]*

Race White 336 (65) [61-69]* 6,101 (58) [57-59]* Hispanic 51 (10) [ 7-13)* 621 ( 6) [5.5-6.5]* Black 129 (25) [21-291* 3,530 (34) [33-351* Asian O ( 0 ) 214 ( 2) [ 0-4 1

Insurance Blue Cross I8 ( 4) [ 2- 6]* Blue Choice 30 I 6) [ 4- 81 Medicare 239 (48) [44-52]* 1,223 (12) [11-13]* Medicaid 1 I2 (22) [18-26]* 2,929 (29) [28-30]* Self-pay 39 ( 8) [ 6-10] 1,126 I l l ) [10-12] Other 62 (12) [ 9-15] 2,950 (29) [11-13]

1,006 (10) [ 9-111* 896 ( 9) [ 7-11]

Social work problem Elderly Alcoholidrug abuse Child abuse Crisis intervention Home services

(nonelderly) Domestic violence Rape Other

271 (49) 32 ( 6) 49 ( 9) 63 ( 1 1 ) 24 ( 4)

10 ( 2 )

104 (19) 1 ( 0.1)

NA NA NA NA NA

NA NA NA

Triage acuity Low acuity (1,2) 84 (18) [15-21]* 7,769 (74) 173-751* High acuity (2 3) 384 (82) [79-85]* 2,665 (26) [25-271*

Median (Interquartile Range)

Age 69 years (31-79)* 26 years (12-42)*

NA = not applicable. *Differences between columns statistically significant p < 0.05.

Financial Data In 1992, at least nine patients were admitted to our

hospital solely for nonacute social reasons. Complete bill- ing information is available for seven of these patients. One patient remains hospitalized and another patient's billing information could not be retrieved. For the seven patients with complete billing information, a total of $58,000 in uncollected charges was absorbed by the hospital ($8,300 per patient). The allowable cost-to-charge ratio in New York State is 0.661, which translates to $5,500 in uncom- pensated care per patient. The lengths of stay for these seven patients were 76, 12, 8, 15, 10, 8, and 29 days. The average length of stay for the hospital was 7.5 days.

I DISCUSSION

The number of patients referred to social workers in this study was far greater than those described in other se- ries,9J3slsJ6 which likely reflects not only the greater needs of these patients but also the greater familiarity of emergency physicians with the activities of social workers in our ED and more extended hours of staffing. Social workers saw an average of 15 patients per day during week days and six patients per day on weekends.

The types of services rendered also were different from those described by others (most commonly psychosocial service to patients with lower illness acuity).9J3J5.16 The social-work staff in this study saw more elderly patients and patients with higher-acuity illnesses. The overrepre- sentation of white patients and underrepresentation of black patients probably reflects the fact that there were many elderly patients, most of whom were white. Many of the elderly patients had multiple problems or were not ambulatory, which put them in higher triage-acuity grades.

Social-work activity is time-consuming, with almost one hour required per patient. In this setting, the prepon- derance of social-work activity was in direct services to patients [(mean number of patients seen per day x mean total time spent in patient-related tasks) + 480 minutes = 63% of the social worker's time per shift]. In addition, social-work staff provided a broad array of services to a diverse population. If social-work staff were not present, many of the service needs would be unmet, be inadequately met, or require additional time from physicians or nurses, thereby, diverting these staff members from medical care activities. Another possible result of inadequate social- work coverage would be physician and nurse staffing increases, if the clinical service time for patients were to remain unchanged.

The greater number of patients during the day shift reflects, to some extent, the lack of night-shift social-work coverage and the consequent backlog of referrals left each morning from the previous night. It is unclear why fewer patients were referred to social workers on the weekend because the total ED volume did not fall to the same extent. It seems unlikely that social problems were more likely to present between 8 AM Monday and 5 PM Friday than on the weekend. The lower numbers are likely due to fewer social- work coverage hours on the weekend and a resultant decrease in referrals (a negative referral bias).

Social workers also provide a valuable educational experience to staff physicians and residents working in the E D . ' ~ s ' ~ Post-release home services for the elderly, alcohol and drug dependency, child abuse, domestic violence, crisis intervention, and rape are some of the emotionally charged areas where social workers have great impact and the most to contribute to physician education. l7

Hospitalization of patients for social reasons, early returns to the ED, prolonged hospital stays due to delays in

Social Work, Wrerzn, Rice 25 1

initiating social service interventions, prolonged ED stays due to the possibility of unsafe release, and dissatisfaction with the plans engendered usually adversely affect the hospital financially. The patient may not have the ability to pay for an admission if payment is denied by governmental or third-party payers. Furthermore, a patient may later need hospitalization that could have been prevented or may bring legal action against the hospital.

In this ED, as in most EDs, social-work services are not directly billed but are part of the indirect costs included in the facility’s Medicare-based rate structures, in which a factor for social services is included. However, changes in the Medicare regulations in 1990 may allow billing for social work under Medicare Part B.18 A Milwaukee hospi- tal has billed for social-work services by increasing the patient’s fee level to the next highest acuity level.19 In the future, social-work billing for certain services should be further explored to help ensure the fiscal viability of a continued social-work presence in the ED.

Even if billing for such services is not feasible, social- work input in the ED is likely to result in cost savings to an institution. Referrals for home services and provision of information and counseling to patients and families may decrease the numbers of patients admitted to hospitals for largely social rather than medical problems. The Cape Cod Hospital in Hyannis, Massachusetts, instituted a program of extended social-work coverage for its ED with the express purpose of decreasing social admissions. Based on their experience, the diversion of 28 potential social ad- missions per year more than paid for the cost of the social- work coverage.20 In our institution, with an average medi- cal census approaching loo%, diversion of as few as 17 social admissions per year (1.5 diversions per month) would pay for the salaries to provide comprehensive social services coverage as described.

Examples of successful social services diversions of nonacute admission include the following cases:

Case 1. An 83-year-old woman with a metastatic brain tumor was brought to the ED because her sole caregiver had been injured in a car accident. Twenty-four-hour aide ser- vice could not be arranged because there was no backup available in case an aide was unable to work a shift. After several hours of intensive social service intervention, nurs- ing home placement directly from the ED was arranged.

Case 2. A 74-year-old woman had been discharged from the hospital 11 days before to her daughter’s home, with services including home health nurse visits, physical therapy, and home health aide care two hours per day, five days per week. She was brought to the ED because the daughter had failed to return home when an aide’s shift ended and she could not be reached. Several concerns about the daughter’s care had arisen during the 11 days. After several hours of intensive social-work intervention involving Adult Protective Ser- vices, a new agreement between the daughter and the home care agency was reached, allowing a safe release from the ED.

I TABLE 2 Distribution of Social-work Interventions Provided to 560 Patients in the ED during a Six-week Period

~

Median Time, No. of Minutes

Service Patients [Interauartile Ranee)

Review/documentation

Information/referral

Risk evaluation

Psychosocial assessment

Case conference

Home care planning

Psychological counseling

Concrete services

Home care follow-up

Interpretation services

Consult only

Miscellaneous

421

238

175

168

1 20

115

74

70

26

22

19

14

20 (10-30)

15 (10-20)

20 (15-30)

20 (15-38)

15 (10-20)

20 (15-30)

30 (20-45)

15 (10-20)

30 (20-30)

23 (20-40)

20 (1 0-30)

20 (15-30)

In addition, release planning and continued social ser- vice follow-up after a high-risk patient leaves the ED may prevent unnecessary return visits and admissions. Pread- mission risk evaluations may decrease in-hospital lengths of stay by early identification and intervention for psycho- social problems that might otherwise delay discharge. This is especialy likely since the preponderance of patients seen by the social workers were elderly or had higher grades of triage acuity. These two areas of social-work involvement may have even greater cost-savings potential than diversion of social admissions. Finally, social workers often help to defuse tension and decrease liability potential from the highly charged encounters that frequently arise among health care providers, the hospital, and patients or their families, all of whom may have different agendas.

Social workers in the ED must have training in caring for patients at both ends of the age spectrum (protective services for children and home services for the elderly), as well as instruction with problems of substance abuse for adults between 25 and 40 years of age. They must be able to deal with people of diverse ethnic backgrounds. Bilingual ability is also useful in many settings. ED social workers must be flexible and able to work under pressure.’ Others have called for more formal training in ED social-work skills and more ongoing educational opportunities. l2

It is impossible to predict the role social-work services will play in the ED with the new health care delivery initiatives now being debated. If the numbers of patients presenting to the ED decrease, it is possible that the need for such services will decline. The increasing numbers of elderly patients with higher-triage-acuity profiles and the push for less expensive outpatient care, however, should

252 ACADEMIC EMERGENCY MEDICINE MAY/JUNE 1994 VOL 1/NO 3

make social-work services even more valuable. In addi- tion, there is no evidence that the cycle of substance abuse and violence in our society is abating. Finally, the acquired immune deficiency syndrome (AIDS) epidemic continues, and provision of complicated home care services is also critical for these patients. Social service interventions are more likely to be relevant in the areas of cost containment in high-volume, urban EDs with overcrowding and high inpatient censuses, as well as those hospitals serving large populations of indigent patients, the elderly, or AIDS patients.

I CONCLUSION

Social-work staff dedicated to the ED provide valuable services not only to the ED patients and staff, but also to the hospital as an institution. Without a comprehensive ED social-work program, these services might be inadequately provided or provided only at a significant cost in physician and nursing time. Cost savings to the institution are possi- ble in several other areas also. In this cost-conscious era, every effort should be made to have social workers as- signed to the ED and to educate physicians and nurses about the nature of the services social workers can provide.

I REFERENCES 1 . Hylton H . Under siege: the embattled trauma care system. Tex Med.

1992: R8:37-47. 2. Shesser R , Kirsch T, Smith J, Hirsch R. An analysis of emergency

department use by patients with minor illness. Ann Emerg Med. 1991: 20:743-8.

3. Sainsburg S. Emergcncy patients who leave without being seen: are urgently i l l or injured patients leaving without care? Mil Med. 1990; 155:460-4.

4 . Pane GA, Farner MC, Salness KA. Health care access problems of medically indigent emergency department walk-in patients. Anri Emerg Med. 1991; 20:730-3.

5. Allison EJ, Detlart KL. The ultimate safety net (editorial). Ann Emerg Med. 1991; 20:820-I.

6. Clark L. How long are your patients languishing in the ER? Mecl Econ. 1990; 67:36-45.

7. Beck M, Gordon J. A dumping ground for granny. Newsweek. 1991; Dec. 23:64.

8. Clement J, Klingbeil KS. The emergency room. Health SOC. Work.

9 . Farber JM. Emergency department social work: a program descrip.. tion and analysis. SOC. Work Health Care. 1978; 4:7-18.

10. Krell GI. Hospital social work should he more than a 9-to-5 position. Hospitals. 1976; 50:99-104.

11. Bennett MJ. Emergency medical services: the social worker's role. Hospitals. 1973; 47:lll-8.

12. Wilson LB, Simson S, Duncan M, Lloyd M. Emergency services and the elderly: the role of the social worker. Health SOC Work. 1982; 7: 5 9 4 4 .

13. Shepard P, Mayer JB, Ryback R. Improving emergency care for the elderly: social work intervention. J Gerontol SOC Work. 1987; 10:123-40.

14. Soskis CW. Emergency room on weekends: the only game in town. Health SOC Work. 1980; 5:37-43.

5. Healy J. Emergency rooms and psychosocial services. Health Soc Work. 1981; 6:36-43.

6. Groner E. Delivery of clinical social work services in the emergency room: a description of an existing program. SOC Work Health Care.

7. Bergman AS. Emergency room: a role for social workers. Health SOC Work. 1976; 1:33-44.

18. Euster S. Ambulatory billing: producing a program. Newsletter NYC Society of Hospital Social Workers. 1991; 1-2.

19. Ponto JM, Berg W. Social work services in the emergency depart- ment: a cost-benefit analysis of an extended coverage program. Health SOC Work. 1992; 17:66-73.

20. Harrington DV. The ER social worker: cost-effective, crisis-oriented discharge planning-and more. Discharge Planning Update. 1991;

1981; 6~83S-9OS.

1978; 4~19-29.

11 18-1 1.

APPENDIX

Definitions

The following definitions, developed by the social-work divi- sion of University Hospital of the University of Rochester, are provided for the purpose of ensuring that social workers and health care providers use the same terminology.

Patient Problems

Elderly-persons 70 years old or older. Alcohol and drug abuse-persons whose use of alcohol and/or

drugs presents significant problems to themselves and/or others. Child abuse-children less than 18 years old who may have

experienced abuse or neglect. Crisis iiiterverition-families of patients who are seriously i l l ,

in,jured, or dead.

Home services, non elderly-persons less than age 70 years old who require support services after ED release.

Domestic violence and rape - persons who have experienced physical, emotional, or sexual assault in the home or sexual assault outside the home.

Social Work Services

Reviewldocumentation-review of medical chart, writing of chart notes on behalf of a patient.

Injormation/referral -provision of information andlor referral regarding appropriate resources to patients and families.

Risk evaluation-initial contact with a referring staff member and review of the medical record, or interview of the patient or family to determine whether identified patients do in fact need social work intervention or release planning.

~ ~ ~ ~ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

Social Work, Wrenn, Rice 253

Psychosorial assessment-obtaining and evaluating informa- tion about the patient and the family or significant others. It includes chart reviews, interviews with patient/family, and con- tacts with hospital and community agency staff.

Case conference - face-to-face or telephone discussion of a particular patient with one or more community agencies and/or hospital staff members for the purpose of treatment or release planning.

Home-care planning-any activity or set of activities that facilitates the release of the patient from the ED.

Psychosocial counseling (crisis intervention)-counseling and/or treatment for patients, families, and/or significant others to provide emotional support and other services to deal with a patient’s critical condition or death.

Further Thoughts from the Reviewers I The authors provide a retrospective, descriptive as- sessment of a social-work service at one urban teaching hospital ED over a six-week period. The article under- scores the social and societal needs of patients who use U.S. EDs. This presentation provides useful data demon- strating the patient support that social workers in the ED can provide to a hospital operating near total bed capacity. It also suggests, but does not prove, that an ED social-work service may be justified financially.

Given the increasing scrutiny of inpatient hospital care, the shortage of staffed hospital beds in many regions, the increase in societal violence, and the aging of our popula- tion, the incorporation of social-work services into ED

Concrete services- brief intervention to resolve specific prob- lems such as facilitating or obtaining transportation, food, shelter, and clothing, or caring for children while parents are being seen.

Home-care follow-up-monitoring that occurs after ED release to ensure continuity of care and to determine whether the post-ED plan was implemented.

Interpreter services-provision of interpreter services for per- sons who have a hearing impairment or whose primary language is other than English.

Consultation only- situations in which the social worker pro- vides advice about a specific patient to ED or community agency staff without providing direct service to the patient and/or family,

Nonacute social admission- admission to the hospital for a reason not directly related to an acute medical problem.

practice seems inevitable. However, subsequent studies must address the actual savings accrued with such a service and to whom these savings accrue. In the absence of a societal health plan in which all parties share in the cost of care delivery, the cost of this service will remain highly visible to the institution providing the service and the overall savings to society may be irrelevant. That is, while the benefits of such a service will reach beyond the institu- tion providing the service, the current medical-economic climate requires that the institution recoup its expenses directly. The onus is upon emergency physicians and social workers to document cost savings to their own institutions sufficient to justify this service.