This project is the pilot step necessary to develop a system of vaccination pharmacoepidemiology that associates a sophisticated analytical approach, appropriate for this challenge and case series analysis, with the use of linked medicoadministrative data for hospitalisations the reimbursement of healthcare costs. This linkage of medicoadministrative data is new in France. The example of the risk of hospitalisation for febrile convulsions in infants less than two years old following vaccination against diphtheria, tetanus and pertussis is given as the reference association. This choice is justified by the numerous publications over many years on the existence of this association (Miller et al BMJ 1981, Farrington et al Lancet 1995). In addition, as case series analyses have been carried out to assess this risk, the results can be compared directly (Lancet 1995). The study to determine whether the analysis tools are suitable for the data will be tackled at the following levels:

• Validity of the selection of cases from administrative records alone using coding based on the international classification of diseases ICD10
• Validity of the risk assessment. This methodology feasibility project must make it possible to identify the different potential problems. This is a prerequisite necessary for the systematic implementation.

Inclusion Criteria:

- For reimbursement data: All of the data concerning delivery (purchase of the vaccine at a pharmacy, prescription) and vaccination (consultation following the date of purchase) for infants younger than 2 years old, in the years 2007 and 2008.

- For PMSI data: All of the hospitalisations in 2007 and 2008 for febrile convulsions in infants less than 3 years old in hospitals of the following departments: Côte d'Or, Doubs and Bas-Rhin

Exclusion Criteria:

-

Children of less than 3 years hospitalized for febrile convulsions