Le volume de prescription des antidépresseurs en France est significativement plus élevé que dans les pays comparables, et il est en augmentation régulière. Toutefois, les conditions exactes de ces prescriptions sont mal connues, en particulier quant à l'adéquation entre le traitement prescrit et la pathologie sous-jacente. L'enquête présentée, réalisée à partir d'un échantillon de 44 000 personnes âgées de plus de 15 ans représentatives de la population française, a exploré les conditions de prescriptions d'antidépresseur en France. Sur cet échantillon, 3,5 % de la population était sous antidépresseur à l'instant t, avec un sex ratio de 3 femmes pour 1 homme ; parmi cette population, 42 % avaient déjà reçu préalablement un antidépresseur. Parmi les consommateurs d'antidépresseurs, 39 % prenaient un tricyclique, 45 % un inhibiteur spécifique de la recapture de la sérotonine. La très grande majorité des sujets recevaient une monothérapie antidépressive. Six prescriptions sur 10 étaient effectuées par un médecin généraliste, et 9 sur 10 par un médecin libéral. Le motif déclaré de la prescription était dans 57 % des cas une dépression, dans 15 % des cas un état d'anxiété ou de stress. Pour 62 % des sujets traités, il existait un diagnostic, effectué à l'aide du M-CIDI, de trouble thymique, et pour 14 % un trouble anxieux ; seuls 5 % des consommateurs d'antidépresseurs ne présentaient ni symptômes ni syndrome psychiques. Parmi les personnes sous antidépresseur, seules 54 % avaient un diagnostic correspondant aux indications de l'AMM du produit considéré. L'enquête souligne le décalage entre l'AMM et les pratiques des médecins prescripteurs.

The consumption of antidepressant seems to be in France higher than in comparable countries, as well as the overall consumption of healthcare and medications. In Western countries, in recent years, the use of antidepressants has regularly increased, mainly due to the use of serotoninergic antidepressants. In France, in a week, the prevalence of antidepressant use in the overall population increased from 1.7 % in 1992 to 3 % in 1995. This survey addressed the overall population in the form of a representative sample focusing on subjects who indicated, at the time they were consulted, that they were taking an antidepressant. The study aimed to determine the circumstances of prescription : prescriber file, reason for prescription, type of medication prescribed, match between the prescription and the product indications stated in the marketing authorization, prescription duration and reason for discontinuing treatment. Methodology – The first stage consisted in forwarding a letter to a panel of 44 000 subjects aged 15 years or more and representative of the French population. The aim was to achieve a cross-sectional description of the population taking antidepressants. The response rate was 82 % (36 036 subjects). The subjects who stated that they were taking an antidepressant were re-contacted by telephone by an interviewer trained in the use of the Composite International Diagnostic Interview – lifetime (CIDI), exploring depression and anxiety diseases with a view to potential diagnosis as per DSM criteria. Longitudinal follow-up over 8 months from the initial screening was evaluated using a monthly questionnaire on the time course of antidepressant consumption. Results – Out of 20 000 households, comprising 44 000 people aged over 15 years, 1 333 people were taking an antidepressant or had taken one in the previous 4 weeks. The sex ratio of the antidepressant consumers was 3 women to 1 man, amplifying the known sex ratio with respect to depressive disorders. The mean age of the subjects taking an antidepressant at time t was 51 years. Lifestyle and socioprofessional category did not seem to influence antidepressant consumption. Somatic comorbidity was present in 60 % of antidepressant consumers. Among the consumers of antidepressants at time t, 45 % were taking a selective serotonin reuptake inhibitor (SSRI). The two products most widely prescribed in that class were fluoxetine (30 % of the subjects taking an antidepressant at time t) and paroxetine (10 % of the subjects taking an antidepressant at time t). The other SSRIs accounted for the remaining 5 %. Thirty-nine percent of the consumers were taking a tricyclic antidepressant : clomipramine in 16 % of cases, amitriptyline in 14 %, and other tricyclic antidepressants in 9 %. Lastly, 20 % of the consumers were taking an antidepressant that was neither an SSRI nor a tricyclic antidepressant. Only 4 % of the patients were concomitantly taking 2 antidepressants : single-agent therapy is in line with the recommendations of the various expert groups. In the survey, 9 antidepressant prescriptions out of 10 were written by an open-care practitioner, and 1 out of 10 by a hospital physician. For 60 % of the subjects, the antidepressant treatment was prescribed by a general practitioner. General practitioners prescribe less tricyclic antidepressants and more SSRIs than specialists. The main reason for prescription reported by the patient was depression (57 % of cases) ; followed by a state of anxiety or stress (15 % of cases). In 10 % of cases, the consumer stated that the reason for treatment was not psychological. Sixty-two percent of subjects presented with, or had presented with, a mood disorder as per M-CIDI (major depression, mood disorder, or a combination of the two) and 14 % an isolated anxiety disorder. Twenty-five percent of the subjects on antidepressants did not fulfill all the M-CIDI criteria for any diagnosis. Among the people receiving antidepressants, 54 % had a CIDI diagnosis in strict compliance with the marketing authorization indications for the product considered. One quarter (25 %) presented with a diagnosis of a characterized psychiatric disease, outside of the marketing authorization indications for the product taken. This finding reflects misuse or use on the basis of published data not incorporated in the marketing authorization. The dosages were in line with those stated in the marketing authorization for the disease considered in almost 99 % of cases for the subjects on paroxetine and fluoxetine, but for only 22 % of cases for the subjects on tricyclic antidepressants. Tricyclic antidepressants would therefore appear to be frequently inappropriately in terms of proportions that would be ineffective : half of the subjects on clomipramine were taking a dose less than or equal to one third of the minimum recommended dose. Conclusion – This survey shows that the point-prevalence of antidepressants in the global population in France is about 3.5 %. Women consume more antidepressants than men. SSRIs are the most widely prescribed antidepressants. The survey findings point out the discrepancies between official indications, such as the ones issued by the regulatory authorities, and the physicians' prescribing practices.