[The acute episode of fever of unknown origin among children under the age of 3: strategy against uncertainty in outpatient medicine].

In general medicine, a febrile state is involved in one pediatric consultation out of three. Whenever he encounters a child with a fever of 38 degrees C or more without any other symptoms, the stake for the family doctor will be to detect a severe bacterial infection (SBI), which would need to be monitored in a pediatric emergency unit, while in 99% of the cases, it will be a benign pathology only requiring simple measures. Based on a review of recent publications, this summary paper provides a decision-making algorithm on the basis of factors readily available in first line care, such as the child's age, his personal history, environment, the intensity and duration of the fever, the possible presence of dyspnea, the level of parental concern and the physician's clinical impression. The management method we are suggesting could be the starting point first, of a field study, through analyzing the way in which GPs actually deal with this problem, allowing for an improvement of the proposed algorithm ; and secondly, of forward-looking intervention work in collaboration with pediatric emergency services, in order to assess its effectiveness in predicting SBI and its ease of use in first line care. It is commonly accepted that the act of decision-making is part of the GP's daily practice, despite its many uncertainties and work without safeguards. Let us expect, however, that this type of research, which can generate predictive scores adapted to first line care, will be able to reduce (but not eliminate) some hesitation and doubts.