(Translated by the author)
1In western countries, the increase in the primary health costs threatens the financial balance of the organizations paying or refunding these expenses. In this respect, the problem of “high users of primary care services”, i.e. patients who frequently consult their general practitioner (GP), is a serious problem for policy makers. This problem is particularly significant for women who are consistently the highest users of services, using about 50 % more health services than men (e.g. Yishai, 1992). These “high users” generate a high workload and economic burden on the primary health services (Neal, Heywood, Morley et al., 1998), while their relationship with their GP is often unsatisfactory (O'Dowd, 1988), and the very fact that they keep coming back may point to needs that are not met by the care they receive. In addition, the proportion of high users of primary care seems to be increasing (Gill, Dawes, Sharpe & Mayou, 1998). Research is thus urgently needed to understand who these high users are and what the significant factors are which contribute to high rates of utilization of primary care services, in order to improve both the efficacy of health care services, as well as the outcomes of medical care.
2Carrying out this type of research in Israel is of particular interest due to several factors. This country maintains one of the highest rates of primary care visits in the world (Yishai, 1992). At the beginning of the nineties, for example, the annual rate of visits was 73 % higher than the U.S. rate (Cunningham & Cornelius, 1993; Shuval, 1990). Although a certain decline of this rate has been observed in Israel in recent years (Central Bureau of Statistics, 1999), the disparity with rates elsewhere still remains. This utilization disparity can in no way be attributable to a poorer Israeli health status: life expectancy is higher than in the U.S. and infant mortality rates are comparable or better, according to the population groups considered (Feinson & Popper, 1995).
3High use of primary care services is obviously the result of a complex interaction of factors such as the characteristics and expectations of GPs and patients, service organization and of course the severity of the patient's illness (perceived or evaluated) (Hulka & Wheat, 1985). One of the major factors which we hypothesize is a significant contributor to the large number of visits is the existence of undetected or untreated mental health problems, since there is evidence that high users of primary care services have a disproportionate amount of different forms of psychological distress and psychiatric disorders (Smith, Monson & Ray, 1986; Katon, Von Korff, Lin et al., 1990; Callahan, Hui, Nienaber et al., 1994; Simpson, Kazmierczak, Power & Sharp, 1994; Karlsson, Lehtinen & Joukamaa, 1995; Dowrick, Bellón & Gómez, 2000).
4A growing number of studies have been devoted to psychiatric morbidity in the general population. Reported rates vary according to the populations studied and the methodology used, but the general picture is that mental health problems are extremely frequent in the general population. In western countries, they are now among the most frequent diseases, and specialists predict that depression will be the second cause of death by the year 2020 (International meeting held in London in October 1999: “Depression, Economic and Social Timebomb”). In their review of the literature, Wittchen, Essau, von Zerssen et al. (1992) have shown that about one third of the adult population has suffered from a psychiatric disorder at some stage in life (lifetime prevalence rates1 of 28 % to 37 %). In a recent study (Kessler, McGonagle, Zhao et al., 1994) even higher rates were found: close to half of the respondents (48 %) reported a lifetime history of psychiatric disorder, most frequently an affective disorder (17 % of the respondents). Other studies carried out in different countries confirm these high rates. Thus the 6-month prevalence rate2 for any psychiatric diagnosis reaches 17.1 % in Edmonton, Canada (Bland, Newman & Orn, 1988) and 19.1 % in the ECA - Epidemiological Catchment Area - sites in the USA (Regier, Boyd, Burke et al., 1993). The reported yearly rate is 23.5 % (nearly 1 in 4 respondents) in the Netherlands (Bijl & Ravelli, 2000) and 29.5 % in the USA (Kessler, McGonagle, Zhao et al., 1994). These very high rates underline the necessity of providing appropriate mental health care to the population. We will thus now examine how the mental problems are addressed and treated.
5It has been repeatedly shown that most people with current defined mental disorders do not look for any treatment for this type of difficulty. Among those people who do use health services, more than half are not treated in the specialty mental health care sector and receive health services only in primary care settings (Goldberg, Benjamin & Creed, 1987). For example, in the Netherlands, only one third (33.9 %) of the people who reported having had one or more psychiatric disorders in the past year sought some form of help (Bijl & Ravelli, 2000). Primary care was sought most frequently (27.2 % of the respondents), while only 15.3 % used mental health care. The general medical sector should thus have a leading role in detecting and treating persons with psychiatric disorders.
6As documented by studies from different countries, a disproportionate percentage of primary care users suffer from mental health problems, whether or not they are physically ill, with estimates ranging from 15-70 %. An Israeli study based on the GHQ (General Health Questionnaire) found emotional distress among 69 % of the patients in primary care clinics in the town of Beer Sheva (Shiber, Maoz, Antonovsky & Antonovsky, 1990). More recent studies, using diagnostic instruments, revealed that between 20 and 45 % of primary care users are classified as having psychiatric disorders. The World Health Organization collaborative study on “psychological problems in general health care” was carried out in primary care settings at 15 sites in different countries (Sartorius, Ustün, Lecrubier & Wittchen, 1996). Well-defined psychological problems were found to be frequent among the patients in care in all the general health settings examined (median 24.0 %). Among the most common were depression, anxiety, alcohol misuse, somatoform disorders and neurasthenia. Severe mental disorders accounted for 1 % to 5 % of the mental health problems seen at the primary health care level. (World Health Organization, 1990). These disorders are generally stigmatized as “madness”. They include psychotic pathologies (schizophrenia, paranoia, major depression), different types of dementia and other organic brain syndromes. Less severe types of psychological disturbance and mental illness, more common but often less easily recognized, include neuroses, acute emotional stress in response to crises such as bereavement or family disruption, chronic stress arising from long-term social and/or economic difficulties, aberrant behavior resulting from personality disorders, etc… Estimates suggest that this type of problem accounts for 20 % – 40 % of all illnesses treated in general health facilities (World Health Organization, 1990). Depression is the mental health disorder most commonly seen in the primary health care setting. According to a recent study (Betrus, Elmore & Hamilton, 1995), the percentage of people who are depressed but are seeking treatment only for physical disorders in the primary care setting ranges between 12 % and 55 % of all patients. In addition, 9 % of patients utilizing primary care services suffered from a “sub-threshold condition” (functional impairment and clinically significant psychological symptoms that did not meet diagnostic criteria) (Sartorius, Ustün, Lecrubier and Wittchen, 1996).
7As described by the World Health Organization (1990), the primary health care settings usually fail to take into account the impact of emotion and behavior on health. They often do not consider psychological and social problems as legitimate health problems that deserve attention in their own right. Primary medicine, centered on an organicist approach to illness, tends to consider the patient as a collection of organs that may sometimes need repair. As a consequence, GPs generally rely heavily on at times sophisticated technology, and numerous functional explorations, on the basis of the patient’s presenting symptoms. This approach is reinforced by the fact that patients suffering from psychological distress frequently present with physical symptoms such as pain, nausea, sexual difficulties, sleep problems, fatigue, lack of appetite or weight loss. Such complaints thus appear to be a way of expressing psychological distress or psychological disorders. General practitioners often fail to detect them as such and, as a consequence, in an attempt to find physical causes for the symptoms, they prescribe repeated expensive investigations and ineffective and costly medications, so that patients visit many different health facilities in a fruitless search for effective treatment. Scarce health resources are thus frequently wasted, while patients are needlessly exposed to treatments that do nothing to help them and may actually give rise to dependence and further mental and emotional difficulties (World Health Organization, 1990).
8Thus, as many have observed (e.g. Maoz, 1998; Mechanic, 1990; Schulberg & Burns, 1988; Shiber, Maoz, Antonovsky & Antonovsky, 1990), the vast majority of mental health problems brought to primary care physicians remain unrecognized, misdiagnosed or inappropriately treated. A convincing body of research indicates that physicians fail to detect most of the psychiatric problems among their patients, failing to diagnose between 45 % and 90 % of mental illnesses in general (Eisenberg, 1992). In the Beer Sheva study, while 69 % of the patients were classified as suffering from emotional distress, only 31 % were identified as such by the physicians (Shiber, Maoz, Antonovsky & Antonovsky, 1990). When mental health symptoms overlap with medical conditions and somatic symptoms, it is only natural for physicians to focus their attention on the physical problems, but the consequence is that the mental health problems, becoming secondary, remain generally untreated or inappropriately treated. As described by Eisenberg (1992), there appears to be a covert agreement between patients and physicians on the fact that physical complaints are the only legitimate ticket of admission to a doctor’s office. Eisenberg (1992) cites a study in which only 20-30 % of patients with emotional distress, family problems, behavioral problems, or sexual dysfunction allowed themselves to report those experiences to their primary care providers.
9Finally, even when a physician recognizes a patient’s need for psychiatric care, the problem of the quality of the provided care subsists. The vast majority of primary care patients with mental health problems are not referred for psychiatric help. They are treated directly by the general practitioners although their knowledge of common emotional disturbances and their skills to deal with them are often insufficient (Davidson, 1986).
10A number of explanations have been put forward – concern about stigmatizing the patient by labeling him/her as suffering from mental problems (the word “mental” is still stigmatized and misunderstood); inadequacy of GPs' mental health training; or the drawbacks of the schedules for reimbursement of mental health services, which tend to penalize physicians who devote time and attention to the patient’s psychosocial needs.
11In short, despite ongoing acknowledgement of the complexity of these issues, they continue to plague health services. As highlighted in a recent Commonwealth Fund report, “very little research has been done on the efficacy of treatment for mental illness provided in primary care settings. An analysis of the evidence suggests that such treatment is often superficial. About half of those visiting general medical clinicians reported that they did not receive the help they needed with their emotional problems. Primary care physicians tend to rely heavily on psychotropic drug therapies, but often prescribe some of these drugs, such as antidepressants, at dosages below accepted therapeutic levels. They rarely refer cases to the specialty mental health sector, referring as few as 10 % of their depressed patients to specialists”(Glied & Kofman, 1995).
12The study we have begun in Israel addresses the question of the contribution of undetected or untreated mental health problems to high rates of utilization of primary health services. The objective and the methodology of this study are described below.
13The general objective of this research is to attempt to enhance the efficacy of the primary care system in the field of mental health. To this end, we plan to evaluate and compare the prevalence of the main kinds of psychopathology among high users of primary health services and among other patients, and, at the same time, to evaluate the detection rate of these disorders by GPs. For this purpose, we will conduct
mental health assessments of the patients in care in primary services (diagnostic interview of the patients);
a study of their GPs’ recognition and treatment practices concerning psychological disorders (interview with the GPs);
an evaluation of the effect of providing GPs with the above-mentioned psychiatric assessment of their patients, on both their treatment practices and patient utilization rates.
14The specific objectives of the study are the following:
To identify high users of primary care services and compare their characteristics (socio-demographic and clinical) with those of a control group of average users;
To measure the prevalence of mental health problems (single psychiatric diagnoses, multiple psychiatric diagnoses and sub-threshold symptoms) among high and average users of primary care services;
To assess physician detection rates of emotional problems and psychiatric disorders among high and average users;
To document physician treatment and referral practices for high users with psychological distress or psychiatric disorders, as compared to average users;
To examine variations in treatment and referral practices according to patient and provider characteristics (e.g. age, gender, ethnic background, immigration status, and type of medical training);
To assess patient satisfaction with health care received in primary care clinics, and in particular with treatment of mental health problems;
To monitor changes in physician treatment practices and patient utilization patterns for high users for whom our diagnostic interview indicated mental health problems undetected by their physician, after the physician has been informed of their mental health status;
To develop policy recommendations for improving detection and treatment of mental health problems by primary care physicians, especially among high users of their services.
15The major hypotheses of the study are the following:
Significantly more mental health problems, both psychological distress and psychiatric disorders, are found among the highest users of primary care services compared to average users;
Primary care physicians detect only a small percentage of the mental health problems, especially when they co-exist with physical health problems;
Physicians are more likely to recognize mental health problems if they conform to societal expectations (e.g. depression among women and alcoholism among men).
Treatment practices vary according to social characteristics of patients and providers as well as type of medical training of primary care physicians (e.g. general medicine, specialization in family medicine or internal medicine).
High users of primary care services with mental health problems express less satisfaction with these services than high users without mental health problems.
16The study began in November 1999 and will last about three years. The overall research design consists of screening patients who attend one of 8 primary care clinics chosen in 4 geographic areas in Israel: Beer-Sheva, with a lower and middle class population (2 clinics); Rehovot, with a higher socio-economic population (1 clinic); Jaffa, with a high proportion of Arabs (1 clinic); Beit-Shemesh, a development town (1 clinic); Jerusalem, (3 clinics, of which one works in collaboration with a psychiatrist and one serves a mainly religious population).
17The screening process consists of asking all patients who come to the clinics during a selected time period to fill out a brief form with key information about age and recent number of visits to the clinic. This enables us to develop a sample of approximately 500 primary care high users and 500 average users, who have utilized the selected clinics for at least one year, are between ages 25 and 74, and have provided informed written consent to participate in the study.
18Participating patients are interviewed face-to-face in the clinic in private, or, if they request it, at home, with a battery of assessment instruments with previously established reliability and validity, including a mental health diagnostic instrument especially designed for administration by laymen (the CIDI-SF – Kessler, Andrews, Mroczek et al., 1998).
19Following the interviews, medical records will be abstracted in order to determine primary care physicians’ recognition rates and treatment practices. Relevant data will be abstracted from the medical records of all interviewed patients (one-year retrospective) after they have provided informed written consent. Information to be abstracted includes: number and dates of visits; reasons for visits; presenting symptoms; diagnoses; symptoms of psychological distress and/or psychiatric diagnosis; referrals to specialists including mental health specialists; drugs prescribed; other treatment practices; other relevant data including risk factors for mental health problems.
20Subsequently, physicians treating the participating patients will be interviewed face-to-face with a specific focus on assessing their recognition of patients with mental health problems and their treatment practices. These data will be obtained using a specially developed and pre-tested questionnaire that will include the following information: general attitudes toward treatment of patients with mental health problems; criteria for identifying patients with emotional problems or psychiatric illnesses; criteria for prescribing psychotropic drugs and dosage usually prescribed; criteria for referring patients to specialty mental health services. In addition, case vignettes will be utilized with physicians to further clarify diagnosis and treatment practices in response to the presentation of mental health problems by various categories of patients.
21For those patients whose mental health problems were undetected by their primary care physician, the physicians will be provided with our diagnostic findings. Six months after providing this information, we will review the medical records to determine what, if any, changes have occurred in the treatment practices of the physicians and in the utilization patterns of patients.
22The study we are conducting in primary care services is expected to shed light on the phenomenon of patients having frequent recourse to these services while their physical status does not seem to justify this pattern. A better understanding of the high use of services may be reached through crossing different types of data. Data obtained from interviews of the GPs will allow us to gain an understanding of which factors in their practice and professional or institutional attitudes may constitute an obstacle to the recognition and treatment of mental health problems. Interview data from patients of primary care services will allow for the detection of those factors which lead to their recurrent requests for care. Finally, data from medical records will enable us to ascertain the possible changes in treatment practices after physicians are provided with the psychiatric diagnoses established during the study.
23The collection and analysis of this database should contribute to the formulation of recommendations for increasing the efficacy of primary care services, and in particular, for modifying their organization.
24As indicated above, our hypothesis is that there is a significant correlation between the inappropriate or excessive use of primary health services, and the absence or inadequacy of mental health problem detection. If the validity of this hypothesis is demonstrated, it will imply that a higher detection rate and more efficient and effective treatment should reduce service utilization. Our study should thus contribute to the improvement of both the quality of care and the satisfaction of patients and their GPs. Finally, in the economic area, it should lead to a greatly needed decrease in public health costs.