|
|
| |||||||||||||||||||||||||||||||||||||||
Background Physicians are sometimes confronted with patients who gain admission to one hospital after another, sometimes referred to as "wandering patients." Little is known about the presenting symptoms of these patients, their use of hospital resources, or the costs of their medical care. We analyzed the demographic and clinical characteristics of wandering patients served by Department of Veterans Affairs medical centers (VAMCs).
Methods For each patient they admit, all 159 hospitals in the Veterans Affairs medical system submit demographic and diagnostic information to a central data base at the Data Processing Center in Austin, Texas. We searched these records to identify patients who were admitted to four or more VAMCs within each year from fiscal year 1988 through 1992. Patients so identified in any one year were called "wanderers"; those identified in all five years were designated "habitual wanderers."
Results We identified 1013 wanderers in 1988. The number gradually declined each year to 729 in 1992. In 1991 there were 810 wandering patients, who averaged about eight admissions per year and more than 100 days of inpatient care; they accounted for about $26.5 million in costs for inpatient and outpatient care in that year. Only 35 patients wandered in all five years from 1988 through 1992. The most common discharge diagnoses of these 35 men were related to substance abuse (mostly alcoholism) and mental disorders. Their 2268 admissions and 7832 outpatient visits cost an estimated $6.5 million over the five-year period.
Conclusions Patients who are repeatedly admitted to different hospitals -- wandering patients -- accumulate high numbers of admissions, cause diagnostic confusion, and receive uncoordinated care. Because of the complexity of their disorders, such patients require case management on a regional or national basis.
Advances in telecommunications have widened access to the comprehensive administrative data bases of the Department of Veterans Affairs hospitals. This information has been of clinical value, especially in identifying previous hospitalizations not reported by patients. We used the Veterans Affairs data bases to determine the prevalence of such wandering behavior, the clinical characteristics of wandering patients, and the costs associated with their care in the Veterans Affairs health system.
Methods
The Department of Veterans Affairs manages the largest health care system in the United States. Each year about 2.6 million veterans, an estimated 10 percent of the total number, receive care from the 159 Veterans Affairs medical centers (VAMCs), 128 nursing homes, 35 residential care facilities (domiciliary units), and 191 community clinics located throughout the United States, Puerto Rico, and the Philippines.
In 1992, 542,894 veterans accounted for a total of 922,514 admissions to Veterans Affairs hospitals; 97.4 percent of these patients were men, and 64 percent were 55 years of age or older. Nearly 40 percent were veterans of World War II, and 26.6 percent were Vietnam veterans.
Patients
In this study a "wandering patient" was defined as someone with admissions to four or more Veterans Affairs hospitals in one year. "Habitually wandering patients" had admissions to four or more hospitals in every year of the study, from 1988 through 1992. Admission to four hospitals was used as the threshold in order to eliminate interhospital referrals and patient relocations unrelated to wandering behavior.
Sources of Data
Every Veterans Affairs facility must submit extensive information on every contact with a patient to the Veterans Affairs Data Processing Center in Austin, Texas. A master record, the Patient Treatment File (PTF), is maintained in an SAS software data base for every hospital admission; this record includes 67 categories of information (fields), including demographic and identifying variables and data on clinical care. There is a similar data base for Veterans Affairs nursing homes and residential care facilities. A third data base contains information submitted each year on about 21 million outpatient visits.
We searched the hospital PTFs for each fiscal year from 1988 through 1992 for social security numbers associated with discharges from four or more hospitals. We created data files, containing all the variables in each record, on the patients thus identified for each year. We created separate files that included information on patients who had wandered in all five years (habitual wanderers); we added information on any admissions of these patients to nursing homes and residential care facilities for the five-year period in order to obtain a comprehensive profile of their use of Veterans Affairs resources.
The annual Veterans Affairs National Cost Distribution Report was used to determine the costs of providing care to wandering and habitually wandering patients. This report presents total costs for various types of patient care (intensive care units, surgical wards, medical outpatient clinics, and so on) and allocates per diem and per visit costs in these service areas. For 1991 these costs were as follows: $254 per day for acute psychiatric care, $475 per day for care in medical wards, and $1,146 per day for medical intensive care.
To supplement the clinical information available from the PTFs, we retrieved approximately five discharge summaries from different hospitals for each of 10 habitually wandering patients.
Results
Wandering Patients
In fiscal year 1988, there were 1013 patients with a record of admission to four or more VAMCs. The number decreased steadily over the years until 1992, when only 729 patients met our definition of wandering patients. Table 1 shows the number of wandering patients each year in the context of decreasing admissions to the Veterans Affairs medical system as a whole.
|
These 810 patients were responsible for 0.66 percent of the total admissions to Veterans Affairs facilities in 1991 but for 2.8 percent of all admissions for acute psychiatric care. At one hospital they accounted for 3.4 percent of the total admissions. Forty-four hospitals admitted one of these patients an average of once a week. In addition to the 6266 admissions, the 810 wandering patients accumulated 22,600 outpatient visits.
The 810 wandering patients accounted for about $25 million in inpatient care and about $1.5 million in outpatient care in 1991 on the basis of the 1991 National Cost Distribution Report.
Habitual Wanderers
About 20 to 25 percent of the wandering patients in any given year continued their peripatetic behavior for a second year. Only 35 patients (about 3 percent) of the 1013 identified as wandering in 1988 maintained this pattern for five years. All these "habitual wanderers" were men, and they had a mean age of 49 years (range, 32 to 72). They averaged about 13 admissions per year, accumulating 20,796 days of inpatient care over the five years from 1988 through 1992 (an average of 119 days per patient per year) (Table 2). In addition to 2268 admissions, they accounted for 7832 outpatient visits.
|
Analysis of diagnosis-related group (DRG) codes suggested that about half the 2268 admissions of the habitual wanderers were related to substance abuse (Table 3). The most common DRG code (26.5 percent) was the irregular discharge of a substance-abusing patient (code 433).
|
|
Discussion
The habitually wandering patients we identified averaged more than one hospital admission per month for five years. By comparison, Geller,7 in a survey of state mental hospitals, identified 108 chronically mentally ill patients with a lifetime mean of 31 admissions (range, 5 to 121). The mean for our group of habitual wanderers was twice that number (64.8 admissions per patient; range, 27 to 142) for a five-year period. In a study of patients repeatedly admitted to general hospitals for physical symptoms without an organic basis, Fink8 identified 56 subjects with a lifetime median of 22 admissions, equivalent to about one third of the admissions of our habitually wandering group during a five-year period.
Habitual wanderers were in a Veterans Affairs hospital for nearly one third of the five-year period covered by our study. They were admitted to 151 different VAMCs and had a considerable effect on some. Two medium-sized hospitals, one in the East and one in the South, had 69 and 70 admissions, respectively, of habitually wandering patients.
Homelessness has been proposed as an explanation for some so-called revolving-door patients.9,10 However, the high rates of irregular discharges from nursing homes and residential care facilities among the habitual wanderers we identified suggest that homelessness is not an adequate explanation for their transient style. Hospitals may be the preferred living situation for some patients who need structure, support, and basic amenities when faced with social problems,11 but habitual wanderers have a more complex problem of maladjustment characterized by impulsivity, transient styles, substance abuse, defects in reality testing, and character disorders.12,13,14,15,16 It is not surprising, therefore, that they can cause diagnostic confusion.
Diagnostic confusion and rapid readmissions, in turn, suggest ineffective treatment and inappropriate use of hospital resources.17 Unfortunately, quality-assurance screening programs disregard readmissions to other facilities because of logistic problems in tracking. Generally, eligible veterans can seek care in a VAMC without the staff's knowing about their previous admissions.
Small numbers of patients consume large proportions of health care expenditures.8,18 Programs concentrating on these patients can be cost effective even when they lead to only a minor reduction in admissions.15,18,19,20 Attention to wandering patients could reduce unnecessary services while addressing their medically dependent lifestyle. Indeed, the Department of Veterans Affairs has a legal responsibility to provide safe care that is coordinated among their own facilities (Nagan J, Veterans Affairs Western Regional Counsel: personal communication).
Some case-management programs designed for resistant patients have been successful.15,21,22 However, patients who continually move about the country present problems never addressed previously. A single data base for all Veterans Affairs facilities would provide an integrated record for each patient, regardless of the location of the facility where care was sought. Clinicians would benefit from having even simple information about patients' use of resources in different facilities. Such a system is possible for the Department of Veterans Affairs, but the technology is not yet readily available. More important, wandering patients who have clinical needs that are not being met would be better served by regional or national case management.
Source Information
From the Portland Veterans Affairs Medical Center (L.P., J.J.) and the Departments of Psychiatry and Medical Psychology, Oregon Health Sciences University (L.P.), Portland.
Address reprint requests to Dr. Pankratz at the Psychology Service (116B), Veterans Affairs Medical Center, P.O. Box 1034, Portland, OR 97207.
References
| |||||||||||||||||||||||||||||||||||||||
Related Letters:
Wandering Patients in the Veterans Affairs System
Howard M. O., Walker R. D., Suchinsky R. T., Turner T. A., Engel A. G., Finkelstein A. S., Engel P. A., Pankratz L., Jackson J.
Extract |
Full Text
N Engl J Med 1995;
332:1582-1584, Jun 8, 1995.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
Previous
