Abstract 25

Background: The study of infantile mortality in France is essentially effected by reviewing death certificates. We present the results of descriptive epidemiological study of pediatric mortality in a French pediatric hospital, Clocheville Hospital, which includes 9 specialist medical and surgical departments and a handicapped children day unit.

Methods: Retrospective study of medical records.Inclusion criteria: all inpatients who died in the hospital from 1.1.1990 to 31.12.1995. Exclusion criteria: sudden infant death syndrome and infants dead on arrival.

Results: Annual admissions: 8891.66 (SD 541.8)(8396 - 9254). Number of death: 62.0 p.a. (SD 8.12) (50 - 72), i.e. 0.7% of admissions. Sex ratio: 0.56. Neonatal mortality represented 52% of all infant deaths and 75% of mortality from 0 to 1 year. In this population withdrawal or withholding of care represented 60% of death. Distribution by department: PICU 49.5 death p.a. (SD8.2)(79.8%) ; all other departments 16.55 (SD3.1) (20.2%).Distribution in the time: Death occurring early: 50% in the 48 hours after admission ; there was a predominance of cumulated deaths during July and august. Causes of death: respiratory insufficiency in neonates (premature and term), cardiac disorders (particularly neonatal and postoperative) and non-tumoral neurological diseases were the main causes in the PICU, tumoral disease predominated in Oncology, cardiac disorders in generalist Medical Pediatric Units and non-tumoral neurological disease in Neurology. For eleven cases there was one ≪ evitable ≫ death (2.9%).

Comments: These findings show 1) that 1/5 children die in a Unit other than PICU, the main causes suggesting that limitation of care may be involved, attention to infants and parents being given by the usual treatment team and 2) that neonatal mortality is essentially the consequence of a medical decision and more related to potential neurological handicap than to failure of reanimation.