RESULTS

1. Characteristics of the primary-care physicians

Among the 370 primary-care physicians, 20.0% were general practitioners, 48.4% were general internists, 16.2% internists with further training in a subspecialty of internal medicine (cardiology, rhumatology, gastroenterology, etc.), 11.9% pediatricians and 3.5% somatic physicians who, as they did not specify their field, were considered to be primary-care physicians. They were mainly male (77.9%) and the average age was 48 1/2.

Nearly three-quarters of them worked in private practice (71.7%), 22.9% in public service and only 5.4% worked in both private and public sectors.

One fifth of these physicians (20.8%) had acquired knowledge of psychiatric problems through post-graduate training in psychiatry. Furthermore, half of them (50.5%) had taken part in group training to improve their interpersonal relationship skills, Balint groups for example. Thus nearly six out of ten primary-care physicians (57.7%) had some training in the field of psychiatry, acquired through one or both of these means.

Primary-care physicians considered that approximately 10% of their patients, on average, suffered from psychiatric problems, and 35% from existential problems. They mentioned the referral of 13 patients a year, on average, to a psychiatrist (private psychiatrists or public mental health services).

2. The understanding of psychiatry by primary-care physicians

2.1. Three main conceptions of psychiatry

In order to show the different conceptions of psychiatry of primary-care physicians, we undertook a factor analysis on 22 items related to the knowledge and skills in psychiatry needed by any physician. The underlying dimensions of the first three factors retained were a psychosocial, a psychodynamic and a clinical-biological dimension.

The additive index created for each of these dimensions thus corresponded to what might be considered as three different conceptions of psychiatry. These conceptions were rather implicit ones, resulting from the emphasis placed by primary-care physicians on certain aspects of their practice which pertained to psychological problems and psychiatric disorders. But it is important to note that they evoked three of the main theoretical models which are commonly referred to in psychiatry itself.

Table 1: The conceptions of psychiatry of primary-care physicians
(items with factor loadings >= .50)

The psychosocial understanding of psychological disorders (Table 1.1) included the complex problems ranging from those arising from chronic illness, aging, substance abuse, to the somatic effects of stress. These problems, which have social implications for the patients, seemed to create important difficulties in the physician-patient relationship.

Within the psychodynamic conception, normal personality development and disorders in children and adolescents were underlined, in a highly psychological perspective. Psychodynamic, or even psycho-analytical knowledge, or an interest in psychological testing, appeared important (Table 1.2).

Finally, the clinical-biological understanding of psychiatry highlighted the importance of establishing a differential diagnosis and of psychiatric treatment and medication (Table 1.3).

2.2 Factors linked to the different conceptions of psychiatry

As far as the psychosocial conception of psychiatry was concerned, it appeared that primary-care physicians under the age of 50 had a greater psychosocial understanding than those who were older (p <.05). With regard to types of medical specialization, general practitioners favored such a conception more markedly (average of the index : 15.1) than general internists (14.3), and particularly internists with subspecialization (13.3), or pediatricians (12.4) (p.<.001). Physicians who had a certain training in psychiatry (p < .01), especially those who participated in group training in interpersonal relationships (p < .01), emphasized this type of understanding more often than physicians without such training.

When we compared primary-care physicians with a high psychosocial understanding of psychiatry (i.e. having values above the mean on the psychosocial index) to those having a lower one (values lower or around the mean), we observed that the former thought that they had among their patients a higher percentage of people suffering from psychiatric (p < .05) or existential problems (p < .01).

A preference for the psychodynamic conception of psychiatry was principally displayed by pediatricians (9.0). General practitioners (7.5) and general internists (7.1) ranked in the middle of the index range while internists with further specialization very rarely showed this orientation (6.1) (p <.001).

Physicians working in private practice fostered such an orientation more markedly than those working in public or in both private and public sectors (p < .01). Primary-care physicians with an important psychodynamic understanding of psychiatry referred a higher number of their patients to psychiatrists (p < .05).

The clinical-biological understanding of psychiatry was endorsed, as was the psychosocial one, mainly by general practitioners (6.3) and general internists (6.2), less often by internists with subspecialization (5.8) and much less by pediatricians (4.5) (p < .001). Physicians with a stronger orientation towards biological psychiatry thought that a relatively high number of their patients had psychiatric (p < .01) or existential (p < .05) problems.

2.3. Degree of interest in psychiatry

It was possible that some physicians might foster two or even three different conceptions of psychiatry. In order to evaluate the degree of their interest in psychiatry, a variable was created based on a combination of the three indices previously described. The indices were first trichotomized, with values below the mean considered as indicating a low degree of interest in psychiatry, values around the mean medium interest, and values above the mean a high degree of interest. Physicians with high values on at least two indices and without low values on the third were considered to have a high degree of interest in psychiatry. Those with at least two low values and none high were considered to have a low degree of psychiatric interest. The others displayed an average one.

More than a fifth (21.4%) of the primary-care physicians showed a particularly high interest in psychiatry, while more than half of them (53.6%) could be considered to have an average one, and one-quarter (25.0%) a low degree of interest. A certain number of characteristics differentiated those physicians with a high degree of interest in psychiatry from those with little interest. The former, who were more often general practitioners (p < .01), had more training than the latter in the psychiatric aspects of their practice (p < .01). They thought that a higher number of their patients suffered from psychiatric (p < .05) or existential (p < .05) problems.

2.4. The conceptions of psychiatry endorsed by specialists other than primary-care physicians

There were important differences between primary-care and non-primary care physicians such as surgeons, dermatologists, radiologists, gynecologists, etc. (n=184). The support of these specialists for the different conceptions of psychiatry, be they psychosocially (p < .001), psychodynamically (p < .05) or biologically oriented (p < .001), was weaker than that of primary-care physicians. Among them, it was mainly surgeons who contributed to creating these differences. The degree of interest of these specialists in psychiatric matters was lower than that of primary-care physicians: half of them showed little interest in psychiatry, whereas only a quarter of primary-care physicians (p < .001) did. They also considered that a lower percentage of their patients had psychiatric (p < .001) or existential (p < .001) problems and they referred fewer patients to psychiatrists than primary-care physicians (p < .001).

When we compared these specialists only to internists with subspecialization, we observed that some differences remained. Specialized internists more often endorsed a clinical-biological understanding of psychiatry and took into account the psychiatric (p < .05) or existential problems (p < .05) of their patients.