Abstract □ 69

Before the dramatic decrease of SIDS due to « back to sleep » campaigns, a winter peak was very prominent; it still exists but is much lower. This along with the clinical symptoms prior to death and the inflammatory lesions seen at histology all suggest infectious agents might be involved in sudden infant death (SID). We reviewed 57 consecutive SID for whom an autopsy was carried out in our Referral Center between Nov. 94 and Oct. 98. These 57 SID were divided into 40 SIDS cases (1 m to 1 y old; mean age = 3.6m) and 17 non-SIDS cases: - 2 pediatric pathologies - 3 accidental deaths and 2 child abuses - 5 neonates (1 d to 1 m) including a metabolic disease, a cardiopathy and an accidental death - 5 children > 12 m (up to 4 y 9 m) including a Thomsen myotonie. The interval between death and autopsy ranged from 4 h to 4 d. The postmortem protocol included clinical, biological (white blood cell count and C reactive prot.), microbiological (viral and chlamydiae serology, 2 sets of blood, CSF and urine cultures, nasal swabs (IF for RSV, adenovirus, influenza and parainfl. virus, PCR for enterovirus) pharyngeal and tracheal aspirations, pleural or pericardial effusions if present, liver, spleen, kidney, lung, heart, meninges, choroidal plexus and stools cultures) radiological and pathological investigations. The microbiological results were always interpretated with regard to the location from which the microorganism was grown, its pathogenic potential, the child's age, and the usual flora of the site. The criteria used for infection were the same as those for living children. In open sites, bacteria were numbered to rule out commensal flora. For closed sites (organs), infection was diagnosed if the same bacterium was grown in 2 different sites. For blood cultures, the organism had to be isolated in 2 different sets of bottles. Histologic findings of shock (fibrin thrombin in capillaries, megacaryocytes in circulation outside lungs, shock kidney and heart) and signs of acute thymic involution were recorded. Severe gastric aspiration was considered as the direct cause of death only when associated with vital signs to diffuse and numerous alveolar spaces containing food material. The data were assessed at a clinico-pathological meeting. Diagnoses were made when the clinical, the biological, the microbiological and the pathological data were in agreement. Results: microbiology was significant in 45/57 (79%) cases (34/40 (85%) SIDS). The same significant bacterium was grown in 2 different sites 58 times, alone or associated with 1 or 2 other bacteria and/or viruses. One or more recognized as pathogenic bacteria were found with a virus in 10 virus alone was found 3 times. The most frequent bacteria was E. coli (27 times, including 12 times alone). The most frequent virus was enterovirus (8 times). CRP was increased in 10/42 cases (8/32 (25%) SIDS). Signs of acute thymic involution were present in 37/55 cases (25/40 (62.5%) SIDS). Aspiration of gastric content was found in 17/57 cases (13/40 (32.5%) SIDS), and a microorganism was identified in 15/17 cases. One or more signs of shock were found in 33/55 cases (22/39 (56.4%) SIDS), and a microorganism was found in 27/33 (81.8%) cases (20/22 (90.9%) SIDS). Conclusion: these results strongly suggest that postmortem microbiology is essential in SID investigation. One should not be allowed to conclude to an unexplained death if no microbiology was done.