Effect of Case Rehospitalization of Ambulatory James
L. Curtis,
Ernest
J.
Elmer
Struening,
Ann
M.D.
Ph.D.
Millman,
D’Ercole,
Ph.D.
Ph.D.
To explore whether intensive case management would reduce discharged psychiatric inpatients’ rates of rehospitalization and increase their use of outpatient ambulatory care services, the authors assigned 435 patients discharged from a public general hospital in New York City to three types of postdischarge services: multidisciplinaty intensive outreach case management, less intensive Community Support System (CSS) case management services, or routine aftercare but no case management. Patients eligible for CSS were generally more impaired and bad more extensive histories of psychiatric hospitalization. At follow-up 18 to 52 months after entering the study, patients who received intensive outreach case management had twice as many psychiatric bospitalizations as the routine aftercare group and bad a pattern of
Management on and Utilization Care Services utilization of hospital-based ambulatory care services similar to the more impaired CSS group. intensive outreach case management did not appear to stimulate sufficient use of ambulatory care services to reduce patients’ frequency of psychiatric hospitalization.
Hospital Center, 506 nue, New York, New York 10037. Dr. Millman is research coordinator of the department of psychiatry at Harlem Hospital Center. Dr. Struening is director of the epidemiology of mental disorders research department at New York State Psychiatric Institute and associate professor at Columbia University School of Public Health. Dr. D’Ercole is director of research at Manhattan Bowery Corporation in New York City.
Case management was identified in the late 1970s as a key feature of support services for deinstitutionalized psychiatric inpatients (1). In such approaches, an individual case manager or a case management team negotiates, coordinates, and advocates for mental health cane, medical care, housing, income maintenance, and other services needed by such patients (1-4). In 1 978 New York State developed a Community Support System (CSS) program to provide case management and other rehabilitative services for patients suffering from persistent and chronic mental illness (5). Eligibility for these special services was restricted to patients who had experienced at least one psychiatric hospitalization ofat least six months’ duration or three psychiatric hospitalizations of at least ten days each within the previous two years. In 1984-85 about half of the inpatients treated in Harlem Hospital Center’s department ofpsychiatry, a facility ofthe New York City Health and Hospitals Corporation, were eligible to receive 90 days of postdischange case management provided through the CSS program. The other patients treated on the unit had too few hospitalizations to meet the eligibility criteria for that program. The presence of a group of patients not eligible for the CSS program presented an opportunity to compare the effects of an intensive outreach
Hospital
September
Dr.
Curtis
partment
is director of psychiatry
and Community
of the
de-
at Harlem Lenox Aye-
Psychiatry
1992
Vol. 43
No.9
case management program offered by Harlem Hospital Center with the effects of the CSS program and the traditional system of aftercare. Specifically, we wanted to explore in a case-control study whether intensive outreach case management would reduce rates of psychiatric rehospitalization after an index hospitalization and increase patients’ use of outpatient ambulatory care services. These questions had important fiscal implications because an average Medicaid reimbursement for an inpatient day of care was $400 in 1984, compared with $90 for the average outpatient visit. Methods Setting. The department of psychiatry at Harlem Hospital Center is the only full-service psychiatric program for the 250,000 residents of the central Harlem community of New York City. The patient population is predominantly African American and low income. About 700 patients were admitted each year to the 47bed inpatient service during 198485 Their average length of stay was 22 days. More than 95 percent of admissions were involuntary or emergency admissions; that is, patients had been found to be a danger to themselves or others. About 80 percent ofthe patients treated in the inpatient service remained in the hospital’s ambulatory care program after discharge. Sample selection. Patients were approached just before discharge, beginning in April 1984 and ending in September 1985 We excluded patients who did not reside in Manhattan, who were not between age 18 and 54, or who had a sole axis I diagnosis of substance abuse or organic mental disorder. Also excluded were patients whose stay was shorter than .
.
895
Table 1 Characteristics from
of psychiatric patients assigned to three Hospital Center, in percentages of patients
Harlem
types
Intensive
of services
at discharge
Community Support System
Total
(N=
(N=430)
Characteristic
case management (N= 146)
Routine aftercare (N= 143)
Black
89.0
95.7 4.3 52.5 95.7
39. 1
92.3 6.3 58.0 90.8 35.9 44.2
50.8
92.3 6.3 56.7 91.8 33.4 44.6
10.3 21.9
10.6 18.0
5.0 31.6
8.6 23.8
Hispanic
8.3
Female BomintheU.S. Divorced, separated, High school graduate Employed
before
59.6 89.0 widowed
27.4
phone contacts were made or attempted at least once a week, and mailgrams were sent if telephone contact was not made. Extensive records were kept of all visits to the home as well as of patients’ visits to the case management team’s office. Each week the team held a case conference with the director of the department of psychiatry to help resolve clinical and administrative issues. Procedure. Patients were selected for the study between April 1984 and September 1985 and were followed for periods ranging from 35 to
14 1)
37.2
index
hospitalization Livedalone1
52 months,
Diagnosis
&hizophrenia1
Alcohol
or drug
ordependence’ Previously
treated
Compared tients (2=3, to have to have
in the
with
37.8
57.4
44.7
37.7
39.9
27.7
35.1
26.0
26.0
48.1
17.9
at
hospital’ 1
39.0 abuse
patients
Community
in the Support
intensive System
case
management
program
were
and routine significantly
more
aftercare likely
groups, to live
df= 1 , p< .05), to have a diagnosis ofschizophrenia(2= 13.14, df= 1, p< been previously treated at the hospital(X2=20.1), df= 1, p< .001) and were a diagnosis ofsubstance abuse or dependence (X =4.64, df= 1 , p< .05).
paalone
.001), and less likely
depending
on their date
of entry into the study. Because patients were enrolled sequentially into the study groups, the three groups’ average lengths of follow-up were equivalent. Data on the study sample’s baseline demographic and biosocial charactenistics were collected using a structured interview administered by case managers before patients were assigned to one of the three study groups. Data derived from the Shared Medical Systems billing system used in the Health and Hospitals Corporation was used to determine patients’ number of psychiatric nehospitalizations at any hospital in the corporation within 18 months of the patient’s index discharge. Additional records from Harlem Hospital Center were analyzed to determine patients’ use of inpatient services at Harlem Hospital Center from 18 to 52 months after the index hospital discharge. Data on use of ambulatory services at Harlem Hospital Center were collected during a three-month period each summer (June through August) for three years, 1986 through 1988.
seven days, because there was not enough time to screen them for inclusion. All patients gave informed consent to participate in the study. A total of 143 patients admitted during this period met criteria for the CSS program. The remaining eligible patients were randomly assigned to either the intensive outreach case management group (N= 147) or the control group who received routine aftercare (N= 145). Treatment conditions. Patients in the intensive outreach case management group received services from the multidisciplinary intensive case management team based at Han1cm Hospital Center. Patients in the CSS program received a less intensive form of case management provided by a different team. Patients in the control group received routine aftercane but no case management senvices. All patients in the study were referred at discharge to appropriate aftercane treatment programs, and most received at least initial treatment from various divisions of the departments ofpsychiatry within the Health and Hospitals Corporation. The intensive case management
team implemented the discharge treatment plan prescribed for each patient in the intensive outreach case management group and monitored the patients’ health problems, psychopathology, family and housing problems, and use of social services. Patients and inpatient staff were informed that the team’s mission was not to assume direct responsibility for cane but to help the patient enroll in a day hospital program, adult mental health clinic, rehabilitation program, on alcohol treatment program. From April 1984 to June 1988 the team consisted ofone or two psychiatnists as team leaders, a social work supervisor, three mental health aides trained to be case managers, one physician’s assistant supervised by an internist, and one or two drivers. Hours were 9 a.m. to midnight, seven days a week. Each case manager was assigned between 35 and 40 patients. The team made home visits once a week in the first month and monthly visits thereafter ifnecessany, with the aim of enlisting the family’s participation in the patient’s care. Tele-
Results Demographic and biosocial characteristics. Table 1 shows baseline demognaphic, biosocial, and diagnostic data for the 430 patients in the samplc for whom these data were complete. Most patients were African Americans, fluent in English, who were born in the United States. The mean±SD age of the patients in the sample was 35.9±12.1 years. Slightly more patients in the sample were women than men, more than half
896
September
Hospital
1992
Vol.
43
No.9
and
Community
Psychiatry
had never married, almost halfhad at least a high school education, and fewer than 10 percent were employed at the time oftheir index hospitalization. Their mean±SD monthly income from all sources was $349± $369. Patients in the CSS program received a mean of one entitlement check per month, compared with a mean of .7 for each of the other two groups, a significant difference (F= 5.23, df=2,427, p
L. Curtis,
Ernest
J.
Elmer
Struening,
Ann
M.D.
Ph.D.
Millman,
D’Ercole,
Ph.D.
Ph.D.
To explore whether intensive case management would reduce discharged psychiatric inpatients’ rates of rehospitalization and increase their use of outpatient ambulatory care services, the authors assigned 435 patients discharged from a public general hospital in New York City to three types of postdischarge services: multidisciplinaty intensive outreach case management, less intensive Community Support System (CSS) case management services, or routine aftercare but no case management. Patients eligible for CSS were generally more impaired and bad more extensive histories of psychiatric hospitalization. At follow-up 18 to 52 months after entering the study, patients who received intensive outreach case management had twice as many psychiatric bospitalizations as the routine aftercare group and bad a pattern of
Management on and Utilization Care Services utilization of hospital-based ambulatory care services similar to the more impaired CSS group. intensive outreach case management did not appear to stimulate sufficient use of ambulatory care services to reduce patients’ frequency of psychiatric hospitalization.
Hospital Center, 506 nue, New York, New York 10037. Dr. Millman is research coordinator of the department of psychiatry at Harlem Hospital Center. Dr. Struening is director of the epidemiology of mental disorders research department at New York State Psychiatric Institute and associate professor at Columbia University School of Public Health. Dr. D’Ercole is director of research at Manhattan Bowery Corporation in New York City.
Case management was identified in the late 1970s as a key feature of support services for deinstitutionalized psychiatric inpatients (1). In such approaches, an individual case manager or a case management team negotiates, coordinates, and advocates for mental health cane, medical care, housing, income maintenance, and other services needed by such patients (1-4). In 1 978 New York State developed a Community Support System (CSS) program to provide case management and other rehabilitative services for patients suffering from persistent and chronic mental illness (5). Eligibility for these special services was restricted to patients who had experienced at least one psychiatric hospitalization ofat least six months’ duration or three psychiatric hospitalizations of at least ten days each within the previous two years. In 1984-85 about half of the inpatients treated in Harlem Hospital Center’s department ofpsychiatry, a facility ofthe New York City Health and Hospitals Corporation, were eligible to receive 90 days of postdischange case management provided through the CSS program. The other patients treated on the unit had too few hospitalizations to meet the eligibility criteria for that program. The presence of a group of patients not eligible for the CSS program presented an opportunity to compare the effects of an intensive outreach
Hospital
September
Dr.
Curtis
partment
is director of psychiatry
and Community
of the
de-
at Harlem Lenox Aye-
Psychiatry
1992
Vol. 43
No.9
case management program offered by Harlem Hospital Center with the effects of the CSS program and the traditional system of aftercare. Specifically, we wanted to explore in a case-control study whether intensive outreach case management would reduce rates of psychiatric rehospitalization after an index hospitalization and increase patients’ use of outpatient ambulatory care services. These questions had important fiscal implications because an average Medicaid reimbursement for an inpatient day of care was $400 in 1984, compared with $90 for the average outpatient visit. Methods Setting. The department of psychiatry at Harlem Hospital Center is the only full-service psychiatric program for the 250,000 residents of the central Harlem community of New York City. The patient population is predominantly African American and low income. About 700 patients were admitted each year to the 47bed inpatient service during 198485 Their average length of stay was 22 days. More than 95 percent of admissions were involuntary or emergency admissions; that is, patients had been found to be a danger to themselves or others. About 80 percent ofthe patients treated in the inpatient service remained in the hospital’s ambulatory care program after discharge. Sample selection. Patients were approached just before discharge, beginning in April 1984 and ending in September 1985 We excluded patients who did not reside in Manhattan, who were not between age 18 and 54, or who had a sole axis I diagnosis of substance abuse or organic mental disorder. Also excluded were patients whose stay was shorter than .
.
895
Table 1 Characteristics from
of psychiatric patients assigned to three Hospital Center, in percentages of patients
Harlem
types
Intensive
of services
at discharge
Community Support System
Total
(N=
(N=430)
Characteristic
case management (N= 146)
Routine aftercare (N= 143)
Black
89.0
95.7 4.3 52.5 95.7
39. 1
92.3 6.3 58.0 90.8 35.9 44.2
50.8
92.3 6.3 56.7 91.8 33.4 44.6
10.3 21.9
10.6 18.0
5.0 31.6
8.6 23.8
Hispanic
8.3
Female BomintheU.S. Divorced, separated, High school graduate Employed
before
59.6 89.0 widowed
27.4
phone contacts were made or attempted at least once a week, and mailgrams were sent if telephone contact was not made. Extensive records were kept of all visits to the home as well as of patients’ visits to the case management team’s office. Each week the team held a case conference with the director of the department of psychiatry to help resolve clinical and administrative issues. Procedure. Patients were selected for the study between April 1984 and September 1985 and were followed for periods ranging from 35 to
14 1)
37.2
index
hospitalization Livedalone1
52 months,
Diagnosis
&hizophrenia1
Alcohol
or drug
ordependence’ Previously
treated
Compared tients (2=3, to have to have
in the
with
37.8
57.4
44.7
37.7
39.9
27.7
35.1
26.0
26.0
48.1
17.9
at
hospital’ 1
39.0 abuse
patients
Community
in the Support
intensive System
case
management
program
were
and routine significantly
more
aftercare likely
groups, to live
df= 1 , p< .05), to have a diagnosis ofschizophrenia(2= 13.14, df= 1, p< been previously treated at the hospital(X2=20.1), df= 1, p< .001) and were a diagnosis ofsubstance abuse or dependence (X =4.64, df= 1 , p< .05).
paalone
.001), and less likely
depending
on their date
of entry into the study. Because patients were enrolled sequentially into the study groups, the three groups’ average lengths of follow-up were equivalent. Data on the study sample’s baseline demographic and biosocial charactenistics were collected using a structured interview administered by case managers before patients were assigned to one of the three study groups. Data derived from the Shared Medical Systems billing system used in the Health and Hospitals Corporation was used to determine patients’ number of psychiatric nehospitalizations at any hospital in the corporation within 18 months of the patient’s index discharge. Additional records from Harlem Hospital Center were analyzed to determine patients’ use of inpatient services at Harlem Hospital Center from 18 to 52 months after the index hospital discharge. Data on use of ambulatory services at Harlem Hospital Center were collected during a three-month period each summer (June through August) for three years, 1986 through 1988.
seven days, because there was not enough time to screen them for inclusion. All patients gave informed consent to participate in the study. A total of 143 patients admitted during this period met criteria for the CSS program. The remaining eligible patients were randomly assigned to either the intensive outreach case management group (N= 147) or the control group who received routine aftercare (N= 145). Treatment conditions. Patients in the intensive outreach case management group received services from the multidisciplinary intensive case management team based at Han1cm Hospital Center. Patients in the CSS program received a less intensive form of case management provided by a different team. Patients in the control group received routine aftercane but no case management senvices. All patients in the study were referred at discharge to appropriate aftercane treatment programs, and most received at least initial treatment from various divisions of the departments ofpsychiatry within the Health and Hospitals Corporation. The intensive case management
team implemented the discharge treatment plan prescribed for each patient in the intensive outreach case management group and monitored the patients’ health problems, psychopathology, family and housing problems, and use of social services. Patients and inpatient staff were informed that the team’s mission was not to assume direct responsibility for cane but to help the patient enroll in a day hospital program, adult mental health clinic, rehabilitation program, on alcohol treatment program. From April 1984 to June 1988 the team consisted ofone or two psychiatnists as team leaders, a social work supervisor, three mental health aides trained to be case managers, one physician’s assistant supervised by an internist, and one or two drivers. Hours were 9 a.m. to midnight, seven days a week. Each case manager was assigned between 35 and 40 patients. The team made home visits once a week in the first month and monthly visits thereafter ifnecessany, with the aim of enlisting the family’s participation in the patient’s care. Tele-
Results Demographic and biosocial characteristics. Table 1 shows baseline demognaphic, biosocial, and diagnostic data for the 430 patients in the samplc for whom these data were complete. Most patients were African Americans, fluent in English, who were born in the United States. The mean±SD age of the patients in the sample was 35.9±12.1 years. Slightly more patients in the sample were women than men, more than half
896
September
Hospital
1992
Vol.
43
No.9
and
Community
Psychiatry
had never married, almost halfhad at least a high school education, and fewer than 10 percent were employed at the time oftheir index hospitalization. Their mean±SD monthly income from all sources was $349± $369. Patients in the CSS program received a mean of one entitlement check per month, compared with a mean of .7 for each of the other two groups, a significant difference (F= 5.23, df=2,427, p
