[Mortality in schizophrenic patients. 3 years follow-up of a cohort]

Although schizophrenia is not in itself a lethal illness, an overmortality in psychiatric patients in comparison to the reference population has been attested to for a long time. Until the use of antibiotics, this overmortality was mainly due to infectious diseases caused by the close quarters in mental institutions. At the present time this overmortality is mainly due to suicide but also with a noteworthy mortality by certain natural causes such as respiratory diseases and cardio-vascular and cerebro-vascular diseases. Some questions still persist as far as cancer is concerned because it represents the cause of death for which the results are the most contradictory and the most surprising. In France psychiatric case registers do not exist. At this moment, there is no systematic registration which allows us to study the mortality of psychiatric patients. If we want to know about the mortality of the mentally ill, we have to conduct special research. Consequently, a special research project concerning the mortality of schizophrenic patients taken into care in public psychiatric sectors was undertaken in 1993. We chose schizophrenic patients because of the greater reliability of the diagnosis and because they are a population affected by over-mortality, particularly by suicide. The importance of mortality studies remains twofold:--They are a good indicator of the quality of health care policy as is for instance the infant mortality rate which remains one of the best indicators of the quality of maternal and infant health care.--They enable the formulation of research hypotheses to be made if you point out specific causes of death in a sub-group like schizophrenics compared to the general population. METHOLOGY: The setting of the research was the public psychiatric sector. Public psychiatric sector is a geographical catchment area of about 70,000 adult inhabitants. The national territory is divided into 800 psychiatric sectors. Sectors volunteered to participate. A selection at random would have led to too many refusal to be pertinent. The sectors were contacted through associations like the French Epidemiologic Psychiatric Group, professional reviews, and telephone contacts. Sectors volunteered to participate but patients had to be included at random. All the patients seen within three months and who met the following inclusion criteria: schizophrenic (ICD10) seen as in-or-out patients, aged 18-64, residing in France and under care in public psychiatric sectors were eligible. Various observations concerning social and demographic characteristics, behavioural risks, physical health, access to private somatic care and psychotropic medication were been made by means of a questionnaire at the time of inclusion. Because we wanted to compare the access to private somatic care, one of the themes studied, consequently we excluded chronic patients, that is to say patients hospitalized for more than one year without interruption and whose access to private somatic care couldn't be comparable in a pertinent way to the general population access.

every year at the anniversary of the inclusion of the patient we ask the sector if the patient is alive, dead or lost sight of. In the last two cases we send a letter to the City hall of their place of birth (procedure approved by National Commission for Informatics and Freedom) in order to know if they are alive or not and if they are dead to know the date of their death. For patients deceased we check then their causes of death in the French National Register of Deaths. For deceased patients we check also their causes of death near the sector. So we have a twofold means of checking up.

3,470 patients were included. These patients came from 122 sectors (15% of all the sectors) spread out over 73% of the territory. The sample did not present statistical differences in its sex and age distribution compared to a national sample. A longitudinal observation has been going on for thre