pital?" The Warren State Hospital is carrying out studies which will follow the released patients in order to determine readmission rates and many other facts but these data will not be reported here." Some of the results of the study of separation rates from Warren State Hospital are shown in figure 8 (a) and appendix table 7, which present the hospital's experience with all first admissions in the periods 1916-25, 1926-35, 1936-45, and 1946-50 with respect to the probabilities of being retained in first admission status, released for the first time, or dying in the hospital. The proportion of patients retained continuously in the hospital has decreased consistently from the earliest period in the study to the present. For example, among the patients admitted in the period 1916-25, 39 percent had been retained for 1 year as compared to 36 percent in the 1926-35 cohort, 33 percent in the 1936-45 cohort, and 28 percent in the cohort of 1946-50. The decrease in percentage of patients retained in the hospital has been accounted for primarily by the increased proportion of patients being released to the community. In the period 1916-25, only 39 percent of all patients had been released at the end of 1 year as compared to 54 percent within 1 year in the period 1946-50. Indeed, in this latter period, at the end of 1 year more patients were released than were released within 5 years during the former period. In none of the periods studied was the proportion of patients released within 5 years after admission less than 50 percent, a much higher proportion than was believed would be the case. There was no difference in the proportion of patients released among patients admitted in the period 1926-35 and 1936-45. There has been little significant variation among the various admission cohorts in the overall proportion of patients dying in the hospital within a specified period following admission. The reason for this is that, although there has been marked reduction in the proportion of deaths among patients under 65 Israel, R. H., and Johnson, N. A.: Outcome of 4,254 consecutive first admissions of schizophrenia. Am. J. Psychiat. A. (In press.) years of age admitted to the Warren State Hospital, this has been counterbalanced by the high mortality in the increased proportion of patients aged 65 years and over now being Figure 8. Percentage of first admissions retained in the hospital, released, and dying in the hospital within specified periods following admission to Warren State Hospital, Warren, Pa., 1916–50, all mental disorders, both sexes: (a) all ages; (b) patients under 65 years of age at time of admission; (c) patients 65 years and over at time of admission. admitted. This is also shown in figure 8 (b and c) and appendix table 7, which present the experience of all first admissions under 65 years and 65 years and over, respectively. It is difficult to interpret the experience for all admissions because of the wide age range of the patients and the varying kinds of disorders presented by them. We might look, therefore, at the experience of patients classified in three broad groups of admission diagnoses: (a) the functional psychoses, which include schizophrenia, manic-depressive psychoses, involutional psychoses, and paranoid conditions; (b) syphilitic psychoses, which include general paresis and psychoses with other forms of syphilis of the central nervous system; and (c) mental diseases of the senium, which include cerebral arteriosclerotic and senile psychoses. The results are shown in figure 9 and appendix table 8. Figure 9 (a) and appendix table 8 show the experience of the functional psychotics. The proportion of such patients retained continuously in the hospital, particularly after the first 3 months of hospitalization, has decreased consistently in the last 30 years. Of patients admitted in the period 1916-25, 52 percent were still in continuous residence 1 year after admission as compared to 35 percent of those admitted in the period 1946-50. This has been due largely to increases in the proportion of patients released. Comparison of the earliest and latest cohorts shows that, within the first year after admission, the proportion of patients released has increased from 42 percent to 62 percent and the proportion of patients dying has dropped from about 6 percent to 4 percent. Although the experiences of the first admissions of 1916-25 and 1926-35 were practically the same, patients admitted in 1936-45 and 1946-50 experienced greater probabilities of release than those admitted in the two earlier periods. The greatest improvement has occurred in the cohort of patients admitted during 1946-50. Although, for patients with functional psychoses, the probability of death within 5 years following admission has always been relatively low as mortality in the mental hospital goes, the probabilities of death in the cohort of 1946-50 have been generally lower than those of previous cohorts. portion of patients retained in the hospital. This is a result of marked decrease in probabilities of death without corresponding increases in probabilities of release. This brings about an accumulation of patients who have been prevented from dying but who have not made sufficient improvement mentally to be released. If the 1916-25 cohort is compared with the 1946-50 cohort within 1 year after admission, the percentage of patients released has increased from 14 percent to 39 percent and the percentage dying has dropped from 56 percent to 18 percent. This has resulted in an increase from 30 percent to 43 percent in the percentage retained in the hospital. A smaller percentage of patients are now dying within 5 years of admission to the hospital than used to die within 1 year of admission. Figure 9 (c) shows the experience of patients with mental diseases of the senium. This group is characterized by extremely high probabilities of death in the hospital and low probabilities of release. Currently, 36 percent of these patients die and 7 percent are released within 3 months following admission. Within 1 year, these percentages increased to 53 percent dead and only 14 percent released. It is striking that there has been very little difference in this picture over the period of this study. Analysis by broad age groups, under 75 years and 75 years and over, also shows no change. For patients with functional psychoses admitted to the Warren State Hospital during the years 1946-50, these data demonstrate that the probability of release in the first year following admission was considerably in excess of the probability of release for patients with functional psychoses admitted during the preceding 30 years. This finding in itself is extremely important. It provides facts for getting community acceptance of the mental hospital as an 'Unfortunately, mental hospital statistics have not yet been sufficiently standardized to make it possible to compare the experience of Warren State Hospital with that of any other hospital in precisely the same terms. However, such data as are available from other hospital systems suggest that trends similar to those seen at Warren State Hospital are occurring in other places. See, for example, discussion of changes in release rates reported in the Proceedings of the Second Conference of Mental Hospital Administrators and Statisticians, pages 2-4 and 32-35 (23). active, dynamic medical facility and for doing away with the idea that, once a patient is committed to a mental hospital, he is lost to his family and to society forever. At the same time, however, this finding raises many important questions that must be answered to provide a better understanding of the factors that have made it possible to return more patients to the community. For example, to what degree have the following been responsible for this increase in release rates? 1. The more intensive use of the various psychiatric therapies such as electroconvulsive therapy, insulin, group psychotherapy, and so on. 2. The kind of risk being admitted to the hospital now as compared to years ago, that is, the severity of the illness as well as the general physical well-being of the patients being admitted now as compared to those admitted in years gone by. (For example, if current admissions contain larger proportions of better risks than formerly, perhaps the differences in release rates are merely a reflection of this fact.) 3. Changes in the attitude of the staff toward the expected condition of the patients at time of release. 4. Administrative factors, such as availability of medical, nursing, and other staff, and organization within the hospital both on a staff level and a patient level. 5. Changes in the community's attitude and that of the patient's family toward the mental hospital and the mentally ill. Many variables must be considered in determining the effect of a specific therapy or of therapeutic programs on the release rates of mental hospital patients. Therefore, welldesigned experimental studies are needed to evaluate the various therapies (shock, group, psychosurgery, and so on), singly and in combination with each other and with various ancillary programs, before statements can be made about the effectiveness of a method of treatment or a treatment program. Such studies should include carefully defined diagnostic groups of patients, comparable control groups, carefully specified therapeutic plans and staffing patterns, and specific objective criteria for evaluating results of treatment and for determining condition at time of release. The questions being raised about the effect of This is particularly important in psychotherapy since the many statements being made about the promise of the tranquilizing drugs have emphasized that these drugs make patients more accessible to this form of therapy. Staffing Patterns the tranquilizing drugs on the outcome of treatment of the mentally ill emphasize quite sharply the need for clarification of what constitutes psychiatric treatment and of the objectives of treatment within the hospital setting. If hypotheses with respect to the effectiveness of these drugs in accomplishing the goals of hospital treatment are to be tested, it is essential that experiments and studies be devised that will permit assessment of the effectiveness of a treatment method without use of the tranquilizing drugs and of the effectiveness of the same method when the drugs are used. Indeed, careful evaluation of the existing psychiatric therapies becomes all the more essential and urgent. Table 2. Personnel employed full-time, and reported vacancies, by occupation, State hospitals for mental disease: United States, 19511 The fact that the tranquilizing drugs make "the disturbed patient calmer although still clear and he therefore becomes accessible to many forms of psychotherapy: group psychotherapy, individual psychotherapy, any of the things of a rehabilitative or curative nature that go on in a hospital" (2, p. 182) raises some serious questions. Among these are: SOURCE: U. S. National Institute of Mental Health: Patients in mental institutions: 1950 and 1951. ington 25, D. C., U. S. Government Printing Office, 1954. 348 335 13 5, 690 5, 436 254 32, 458 30, 644 1, 814 Wash Table 3. Selected ratios of patients to personnel employed full time in State hospitals for the prolonged care of psychiatric patients: United States, 1941–51 * Excludes superintendents, assistant superintendents, and medical interns. SOURCE: U. S. National Institute of Mental Health: Patients in Mental Institutions: 1950 and 1951. U. S. Government Printing Office, 1954. Washington 25, D. C. How should staffing patterns be changed? How will existing staffs have to be retrained? How many additional, or perhaps fewer, personnel will be required to do the job that will need to be done in the new milieu being created by the tranquilizing drugs? A large number of persons with a variety of training, skills, and experience are required to provide for the treatment, care, and housing of the hospitalized mentally ill. As of the end of 1951, more than 110,000 persons were employed to take care of the needs of the 500,000 patients in the long-term State mental hospitals, a ratio of 4.5 patients per employee. The distribution of these employees by occupational class is given in table 2. The data available on a national level concerning mental hospital employees are quite limited. For the most part, they consist of ratios of patients to personnel and some gross rates of turnover of various categories of personnel. Nevertheless, these data indicate that there a lack of qualified personnel, particularly psychiatrists, other physicians, nurses, attendants, psychiatric social workers, and psychologists. Table 3 shows the trend in ratios of patients to selected categories of personnel in the years 1941-51 in the long-term State hospitals. The In addition to inadequate numbers, the turnover rate for personnel is quite high. Council of State Governments studied these rates of turnover during 1949 as measured by the ratio of separations within a given personnel category to the maximum number of positions filled during the year. The summary table from the council's report (16) is included as table 4. The council commented as follows: The turnover of personnel in State hospitals in 1949 was relatively high in comparison with turnover rates in other types of public employment. The average annual rate of turnover for physicians was 28 percent; for graduate nurses, 34 percent; for attendants, 55 percent; and for social workers, 17 percent. This situation is particularly unfortunate from the standpoint of the |