Assessing the effectiveness of treatment is one of the main concerns of any medical process. The different ways proposed for assuming the responsibility of drug addicts and their efficacy are greatly heterogeneous since drug-abusing takes on diverse forms. Thus, in order to closely target the request of drug addicts and adjust their follow-up, we undertook to study prospectively, for 3 months, a population of drug addicts taking medical advice for the first time at the Cassini center in Paris, with the assumption that some predictors may forecast outcomes. Method : data were obtained at the admission with a structured interview about socio-economic and demographic status, psychiatric disorders (assessed clinically according to DSM III-R and with HAD and MADRS scales), substance use and prior treatment history, environmental data las well as familial substance use or support lending). Medical and paramedical referents have been interviewed after their first contact with the patient about his expectations and his motivation. Familial attendance at this first contact was noted as well as ifs implication in the programm. At the end of the study we noted lengh of stay, regularity of follow up and clinical changes with a last interview of the staff. Results : half the time, patients' follow up doesn't last a month, drug abuse doesn't change in 6 out of 10 cases, and we only note 14 % of durable abstinence. Polydrug abuse lover 80 %) is not linked, here, with pejorative outcomes, in opposition to the usual literature. Heroin is the main substance used by our population (over 80 %), other opiates, sedatives and alcohol are associated by more than 30 % of these patients; cocain is associated in a quarter of the cases. More than 10 % of the patients are concerned by ecstasy and LSD. Cannabis use is common, Medical complaint (mainly viral diseases) at the begining of the programm, concerns one of two patients. Only a few are initialy known as being HIV positive, suggesting a great lack of information. Over forty percent of the patients are given a DSM III-R diagnosis at the end of the first medical advice, when a doubt subsists for a third of the other patients. Major depressive disorder for the first axis and borderline personality disorder for the second axis are the main disorders we founded. We also noted a large ratio (n = 13 5 %) of schizophrenic disorders. Univariate analysis: length and regularity of the programms are key factors of their efficacy. A long follow-up is also required to improve patients' socio-economical status. initial psychiatric disorders are linked, in our study as well as in literature, with longer stays in therapeutic programms. By revenge, psychiatric disorders at the third month lover 10 %) are linked with poorer outcomes. We noted with interest that in our sample, neither imprisonment in the past (over 40 %, but we noted several imprisonments in a case out of two), or intra veinous route at any moment of the patient's life time (40 %), or else a programm caused by a court (a quarter of the patients) are of wrong prognosis. Relatives' implication in the programm is linked with favorable outcomes. Multivariate analysis draws 3 independant clusters about the length of stay. One concerns patient's motivation as assessed by medical staff. An other one concerns patient's relatives' implication in the care. A third one is about the begining of the treatment. an initial medical prescrition and a psychological help are linked with favorable outcomes. About the efficacy, multivariate analysis isolates 4 independant clusters. Prior drug abuse programms tone out of three patients) are associated with poorer outcomes, when, by revenge, familial relationship initialy seen by the patient as (very) satisfaying, patient's motivation, and, again, an initial medical prescription are linked with better outcomes. The study of those of the patients whose programm lasted more then 3 months but without any appreciable benefit shows that a long follow-up is successfull when it is regular, when it provides a socio-economical status improvement and when the patient is given access to insight For these patients, the told) age is associated with better outcomes. By revenge, such a 3 months follow-up is not able to reduce drug abuse when a psychiatric disorder exists at the third month. Patients whose treatment was refered by a court don't differ from the others: their length of stay and outcomes are the same. Discussion : our study confirms our initial hypothesis according to witch subgroups in our population of drug abusers should be isolated and that some predictors of outcomes should be described. Three points seem important to be discussed. First, a medical presciption appears to be important to initiate the relationship between the patient and his practicioner. We have never see any report about this particular point. Second, our study proves the importance of the patient's relatives' implication in the programm. Finaly, a programm ordered by a court is not pejorative per se. According to the literature, the other results we obtained corroborate some factors detected by other studies, and emphasize the priority of subjective contingents, especially psychopathological ones, over socio-demographical elements. The main limit of this study holds in the fact that our sample only concerns 78 % of the population we could include in the study, according to the well known difficulties of such studies. It should be of interest to complete this study, especially to confirm these predictors on a longer term, in other centers and cities. Conclusion : according to the literature, this study accounts for the necessary global and sustained work with drug abusers. We have to answer to these patients' multiple requests, that don't just concern drug abuse, but also land sometimes merely) psychiatric suffering, social isolation, or medical disease.
