Background
The Israeli psychiatric case register, maintained since 1950, was enacted into law in the Treatment of Mentally Ill Act of 1955 and reaffirmed in the updated 1991 version of the law. Institutions with psychiatric admissions are required to file a report of each admission with the Ministry of Health. Since 1950 the names of about 130,000 patients have been entered into the psychiatric case register. About 93,000 of those patients are still alive. Approximately 16,000 hospitalizations are recorded annually, including some 4,000 first-time admissions whose names are added to the register.
As with most nationwide registers (
2), the Israeli psychiatric case register covers primarily inpatient services, including information pertaining to the hospitalizations of adults and children in psychiatric hospitals and in psychiatric departments of general hospitals. Overall, 34 institutions are required by law to report information to the psychiatric case register. Ambulatory visits in any setting are not included, although day hospitalizations are.
Israel's psychiatric case register, which covers a population of close to six million, is perhaps the world's largest. The only other nations that maintain a nationwide psychiatric case register are Denmark, with a population 5.3 million (
3), and New Zealand, with a population of 3.6 million (
4). Norway, with a population of 8 million, founded a psychiatric case register in 1916 (
5). However, it was abolished in 1988 because of public concerns about confidentiality (
3).
The special advantage of a nationwide psychiatric case register is that it can eliminate the confounding effect of migration on the calculation of changes in an area's rates of hospitalization over time (
6,
7). For example, areas with a declining population may be expected over time to have an increasingly destitute population with a higher rate of long-stay inpatients, although this rate may actually be the artifact of the contracting general population. A nationwide register can avoid this pitfall.
A shortcoming of nationwide registers is that they are essentially limited to inpatient services. The possibility of maintaining records of a wider range of services on such a broad scale without compromising the reliability of the information has usually proven too daunting a task (
2). In Israel, a record of patient contacts at government psychiatric outpatient clinics is also maintained and has been computerized in recent years (
8), but this information is not centrally pooled, for reasons of maintaining confidentiality. On the other hand, many other countries maintain regional psychiatric case registers that cover both inpatient and various outpatient services, providing a fuller picture of a patient's treatment history as he or she moves through both the hospital and the community. Eight such registers exist in England alone, in areas with populations varying from 100,000 to 500,000 (
9). Twelve other European countries (
10), as well as parts of the United States (
11) and Australia (
12), maintain regional psychiatric case registers.
In Israel as elsewhere, the data included in the standardized report of hospitalization submitted by the admitting institution has been altered over the years to reflect changes in psychiatric practices. For example, data from day hospitalization units were first included in the register in 1980. The box above summarizes the types of data currently included in the Israeli register. A total of 26 data fields are currently included, which is similar to the number of fields in most psychiatric case registers (
13).
Accuracy of the demographic data is enhanced by comparing the information submitted by the hospital about the patient with the data available through the population files of the Ministry of the Interior. The psychiatric case register is also regularly updated to include data on patients' deaths, even when they occur many years after the last discharge.
Although the types of data collected by different psychiatric case registers vary, they all include data on psychiatric diagnoses based on one of the various classification systems (
14). Since 1997 the Israeli psychiatric case register has used the
ICD-10 for diagnostic classification. The diagnosis is assigned clinically, usually by the admitting physician, and may be altered at discharge by the treating psychiatrist. The diagnoses are rarely determined by structured interview.
A study undertaken to determine the reliability of these diagnoses found that nearly 60 percent of diagnoses recorded in the psychiatric case register did not change over time (
15). Among those that did change, 90 percent of the changes took place during the first admission and not subsequently.
Research elsewhere on reliability of case register diagnoses has been mixed. In Ireland a comparison of diagnoses in the psychiatric case register with those derived from an interview using the Present State Examination showed a high concordance (
16). A World Health Organization collaborative study in the United Kingdom showed an 85 percent concordance between psychiatric case register and
ICD-9 diagnoses (
17). A Danish study suggested that the diagnosis of schizophrenia is more likely to be given after the first admission (
3). A subsequent comparative investigation showed that patients were more likely to be diagnosed as having schizophrenia during their first admission in Mannheim, Germany, than in Denmark (
18). One may conclude that despite advances in the reliability of recorded diagnoses, it remains a potential weak spot in psychiatric case registers (
19).
Problems with the Israeli case register
Despite the many uses to which the Israeli psychiatric case register has been put, certain problems remain. If they were rectified, the value of the case register would be enhanced.
First, the case register primarily includes data only on inpatient services. Several reasons exist for this limitation. For example, the register was established at a time when community services were not developed. In addition, at least part of its rationale included the need to prevent certain "undesirables" from serving in the security forces (
39). Furthermore, because the register is nationwide, the collection of data from the full range of psychiatric services is a potentially overwhelming task (
2).
It is time to reconsider the case register's limitation to services offered in inpatient settings. The government-operated community clinics have computerized databases (
8). This huge body of information is not collected centrally, due to a well-founded sensitivity to the considerable risk of a breach in medical confidentiality. Nevertheless, separate local outpatient databases can be combined with the inpatient psychiatric case register to provide a more comprehensive picture of the mental health system, as has been done successfully elsewhere (
19). Confidentiality can be protected if the information is transmitted to a central pool without including identifying data such as name and identification number. Recently devised computerized records from community-based vocational and residential settings could also be included. The need for a rational policy of deinstitutionalization demands that such information be made available.
A second problem is that the psychiatric case register has not been mined for its research potential. Studies about patients' mortality, fertility, and season of birth have not been carried out. A major advance in this area would entail linking the register with other databases. For example, in Norway and Denmark linkages between the psychiatric case register and the register of twins were used to evaluate the genetic contribution to schizophrenia by comparing concordance between monozygotic and dizygotic twins (
5,
13). Linking the Israeli psychiatric case register to other databases would similarly enlarge the amount of information to be culled for studies of season of birth, obstetrical complications, mortality patterns, and migration.
A third problem is that the psychiatric case register's potential for discerning trends in the inpatient population—such as the degree of chronicity, the characteristics of new chronic patients, and mortality patterns—has not been realized. Linkage to other databases would add to this potential. However, the psychiatric case register may be most valuable in defining research questions or identifying the cohort to be followed. The actual research must then be done in the field and not in front of the computer.
For example, as noted earlier, case register data was used to determine that mental illness—especially schizophrenia—led to reduced socioeconomic status, rather than being caused by it (
34,
35). To confirm these results, a field study involving interviews with 4,914 adults born in Israel was undertaken. The findings for schizophrenia were confirmed, yet findings for depression among women and substance abuse disorders among men suggested that social causation was more noteworthy (
41).
Similarly the psychiatric case register should be exploited more often as a basis for further field studies, as in the two examples noted: a study of case management policy, intended to follow a cohort of revolving-door patients identified through the psychiatric case register, and a study of brief admissions, which vary greatly from hospital to hospital and must be assessed on site. If research strategies involving the register are carried out more frequently, more questions of policy, quality assurance, and research will be resolved.
A fourth concern is the need to add new types of data to the register. The types of information included are periodically reviewed and updated, and the demands of deinstitutionalization suggest the need to collect additional types of data. For example, is the patient being admitted from a hostel, a sheltered apartment, or his or her own home? Does the patient work in competitive employment, in a sheltered workshop, or not at all? Moreover, data on costs might help in assessing the economics of hospitalization, especially if the differential scale for remuneration of hospitals takes effect (
42).