RESUMO
OBJECTIVES: To describe asthma deaths and to identify the principal short term risk factors in a population followed by twelve thoracic physicians in the South West of France between 1989 and 1995. METHODS: A multicentre clinical study in 312 pure asthmatics recruited freely and reviewed on average 37.4 months later by the same specialist. The attribution of deaths due to asthma was discussed case by case. Predicted risk factors for survival have been researched from variables on the identify card (T0) and analyses using Cox's model. RESULTS: Twenty one subjects died, 16 were due to asthma. This was an annual mortality due to asthma of 1.6%. Seven deaths occurred at home (an acute crisis was the most common) and nine in hospital (respiratory prolonged distress). The principal variables linked to poor survival were: advanced years, asthma with continuous dyspnoea, previous stay in intensive care unit, taking inhaled beta 2 mimetics continuously after inclusion, decreasing FEV1 (P < 0.001). Other variables had a weaker link: duration of asthma, exacerbation of asthma at T0 (P = 0.01), social and psychological problems (P = 0.05), severity of disease at T0 (1992 consensus) were linked to asthma mortality: 75% of the deaths concerned severe asthmatics but 25% occurred in an unforeseen fashion in moderate asthmatics. The presence of personal allergy and of rhinitis was linked to better survival independent of age and severity of asthma. CONCLUSION: This study has shown that it was possible to determine predicted risk factors for asthma deaths in the short term at the time of consultation with an asthmatic and so, to identify the most threatened subjects. The institution of assistance programs could be beneficial as a result of the knowledge of these risk factors.