mental diseases of the senium continue to rise in frequency until the end of the life span.

The characteristics of the resident patientsthat is, the population resident in State mental hospitals on any one day-are considerably different from those of the first admissions.

Although data on resident patients by age, sex, diagnosis, and length of stay are not available for the United States as a whole for 1950, such data are available for selected States.2

The resident patient curve specific for age and diagnosis yields interesting information with respect to age differences in resident patients with specified types of mental disorders (fig. 5 and appendix table 5). These differences might be highlighted by considering two important diagnostic groups, the schizophrenics and patients with mental diseases of the senium. The schizophrenics constitute a high proportion of resident patients in all age groups whereas

'California, Louisiana, Michigan, Nebraska, Ohio, Pennsylvania, and Virginia, 1950.

patients with mental diseases of the senium become a major problem only in the age group 65 years and over.

Although admissions of senile cases have increased greatly in the last decade, the resident population of mental hospitals consists largely of a slowly accumulated core of schizophrenic patients, who are admitted during youth or early maturity and stay, in many cases, for the rest of their lives. The turnover of senile cases is very rapid because of their high death rate. Thus, the median duration of hospitalization for resident schizophrenics, who constitute about 46 percent of the resident population and 24 percent of first admissions, was 10.5 years. For resident patients with mental diseases of the senium, who constitute about 12 percent of the resident population and 26 percent of first admissions, the median duration of hospitalization was 2.4 years.

Although mental hospitals provide care and treatment for persons afflicted primarily with Figure 5. Resident patient rates per 100,000 civilian population for selected diagnoses, by age: State hospitals for mental disease, California, Louisiana, Michigan, Nebraska, Ohio, Pennsylvania, Virginia, 1950.

RATE PER 100,000 CIVILIAN POPULATION

1,000

1,000

UNDER 15- 20-25-30- 35-40-45- 50- 55-60- 65-70 AND 15 19 24 29 34 39 44 49 54 59 64 69 AGE IN YEARS

the psychoses and for a smaller proportion of patients with other disorders, it should be emphasized that the patients constitute quite a heterogeneous population group of mentally ill persons, many of whom have superimposed or concurrent physical illnesses. Of the patients under treatment, some primarily the new admissions are in the acute stages of illness, whereas others are chronic, long-term residents. Some are quite young and others are quite old, and their diagnoses are varied. Although all socioeconomic classes are represented, the great majority of patients in the public mental hospitals are drawn from the lower strata of society.

Because of the magnitude of the problem, hospital administrators and public health, welfare, and governmental officials at the Federal, State, and local levels are interested in finding some way (a) to reduce first admission rates to these hospitals or (b) to effect a higher turnover of patients, that is, to increase release rates and to keep readmission rates low, thus eventually reducing the size of these populations. Already mental hospital administrators have mentioned that they have noticed either a stabilization of resident populations or a tendency for some populations to decrease. A recent letter from the Council of State Governments also raises this point. Although some of these trends had started some years ago, it is interesting that these questions have been raised more frequently since the tranquilizing drugs have come on the scene.

3

Although the tranquilizing drugs do possess some of the necessary properties of an agent that could prevent admissions to mental hospitals by making it possible to treat more patients on an outpatient basis, much more information is needed about the processes operating in society that lead to hospitalization before one can conclude that any significant proportion of a reduction in admissions to mental hospitals can be attributed to the tranquilizing drugs. The fact that reserpine can produce severe mental disorders in patients in whom these disorders were not manifest prior to the use of the drug also indicates that

3 Communication from Sidney Spector, director, Interstate Clearing House on Mental Health, Council of State Governments, 1313 East Sixtieth Street, Chicago 37, Ill.

it has a property of producing a condition which may require hospitalization. In addition, much more information is needed about the factors that contribute to improvement in patients and that lead to release from the hospital before any major portion of these differences can be attributed to the tranquilizing drugs.

The explanation of trends in admission, release, death, and resident patient rates of mental hospital populations and of differences that may be found between corresponding rates in different States, regions, and countries is no simple matter. A mental hospital is a complex population unit to study. The composition of its patient population is a resultant of medical, social, economic, and administrative factors which have determined (a) current and past rates of first admissions to the hospital, (b) current and past rates at which patients are released to the community or die, and (c) current and past rates at which patients are readmitted to the hospital (19). Furthermore, a great variety of factors must be taken into account in order to explain adequately the observed differences. The etiologies of the major mental disorders are unknown, diagnostic methods are not fully standardized, and the relative efficacy of the therapies used to treat the mentally ill lacks careful statistical evaluation.

Let us consider some of the questions raised by studies of indices of the type demonstrated.

First Admission Rates

First admission rates, long regarded as an incidence index for the psychoses, have been used to answer questions on whether there has been an increase or decrease in the incidence of specific disorders, that is, the rate at which new cases develop (20, 21). Such studies, however, no matter how carefully done, suffer from the limitation that the relationship existing between the number of persons hospitalized for a given disorder and the number

Kramer, M.: Some considerations of the population dynamics of mental hospitals with a discussion of research needed to evaluate the accomplishments of such hospitals. To be published in the Proceedings of the Research Conference on Socio-Environmental Aspects of Patient Treatment in Mental Hospitals, December 13-16, 1955, Boston Psychopathic Hospital, Boston, Mass.

of persons in the population who have the same disorder but who never reach a mental hospital is not known.

Hospitalization rates are a resultant of the incidence of mental disorder and of a series of factors that determine the number of persons who are admitted eventually to mental hospitals. These factors include availability of mental hospital beds; availability and usage of other community resources for diagnosis and treatment of mental disorder (for example,

general hospitals with psychiatric treatment services, psychiatric clinics, and private psychiatrists); and public attitudes toward hospitalization. Therefore, to understand more fully the distribution and course of mental illness in the population, it is necessary to study rates of hospitalization in mental hospitals in relation to these factors.

The following example illustrates why studies of the way the mental hospital is used by the community of which it is a part or, to borrow

SOURCE: Patients in mental institutions.

tute of Mental Health, 1947-51.

25489 0-58-16

a term from Clausen (22), studies of "paths to the mental hospital" and, conversely, studies of "barriers" hospital administrators place between the hospital and the community are essential to our understanding of the natural history of mental hospital populations.

Consider the changes in the first admission rate to State mental hospitals of the Nation for persons 65 years and over between 1940 and 1950. In this 10-year interval the rate has increased from 170.4 per 100,000 population

65 years of age and over to 204.9, an increase of 20 percent. The actual level of this rate in the different States in 1950 showed striking variation. The rate varied from 416 per 100,000 in New Hampshire to 74 per 100,000 in Kansas, with a median rate of about 200 (fig. 6 and appendix table 6).

The percentage increase in the first admission rate in the different State hospital systems of the Nation also showed striking variation (fig. 7 and appendix table 6).

SOURCE: Patients in mental institutions.

tute of Mental Health, 1947-51.

Of the 41 States included in this study, 30 showed increases in this rate, 1 showed no increase, and 10 showed a sizeable decrease. The percentage increase varied widely, from 174 percent in Texas to 4 percent in Florida to no increase in Virginia and to a 41-percent decrease in Maryland.

What do these differences mean? Do they reflect true geographic differences in the incidence rate of psychopathology in this age group or do they represent differences in the way the different States handle the problems of psychopathology of the aged? Similar geographic differences can be shown in the age-specific first admission rates to mental hospitals in every age group. It is essential that we undertake studies to explain such differences if we are to understand who gets into the mental hospital, why he gets there, the way he gets there, and what relationship these events may have to his subsequent hospital stay and possible recovery. To do this requires the development of practical casefinding methods and standardized diagnostic techniques for detecting and classifying various psychiatric disorders in the general population. Much research is needed to solve this difficult problem.

It should be clear, therefore, that much careful epidemiological and social science research is needed before it can be determined what part the tranquilizing drugs may be playing in the reduction of admissions to mental hospitals.

Separation Rates

The advent of the tranquilizing drugs has also suggested the possibility that agents are now available that may make it possible to return more patients to the community. The interpretation of release rates from mental hospital populations and the many factors that may be responsible for them is quite a complicated problem. The results of a study carried out jointly by the National Institute of Mental Health and the Warren State Hospital, Warren, Pa. (19), emphasize not only the complex nature of the problem but also the fact that even prior to the advent of the tranquilizing drugs, striking changes had been occurring in the release rates of patients from mental hospitals.

"What has been the trend during the period 1916–50 in the probabilities of separation from the hospital, either by release to the community or by death, within specified periods following first admission for patients of specific age, sex and diagnosis?"

For purposes of this study, patients were followed from the date of their first admission to the hospital to the date of the first significant movement out of the hospital. The "first significant movement out of the hospital" is defined as the date of first release to the community on direct discharge or the date of discharge to convalescent care, whichever came first, or the date of death in the hospital. Patients admitted prior to 1946 were followed until January 1, 1951. Patients admitted in the period 1946-50 were followed until January 1, 1954, so that each member of this cohort could be followed for at least 3 years. Essentially all of the patients admitted to the Warren State Hospital are committed. Voluntary admissions have been so infrequent as to be. negligible. For example, in the year 1950, only 3 of the 709 first admissions were voluntary.

In this type of study, patients who have been released and then readmitted to the hospital do not reenter the study group. This is done not because of any disinterest in finding out what happens to discharged patients but because it seems logical to determine first, "What are the chances of first release to the community?" and then "What are the chances of released patients being readmitted to the hos

A cohort is a group of persons, each of whom has some common characteristic. It is a term commonly found in the literature dealing with statistical methodology of followup studies. For example, in this study, cohorts of first admissions are identified by such common characteristics as the year of admission to the hospital, age, sex, and diagnosis of a patient.