[An epidemiological survey of sudden infant death syndrome in Tokyo]

A complete autopsy is essential for an accurate diagnosis of sudden infant death syndrome (SIDS), whereas the autopsy rate is low in Japan. This study investigated the incidence of SIDS in Tokyo, where the autopsy rate for SIDS was the highest in Japan.

We selected death certificates of infants aged less than 2 years during 1990 to 1992 in Tokyo, whose cause of death was suspected to be SIDS, and classified those possibilities of SIDS according to death descriptions.

The incidences of SIDS by strict definition and broad definition were 0.23 and 0.32 per 1,000 live births, respectively. Not a few death certificates described causes of death and death scene at some length. If including cases of suspected SIDS, the incidence came to 0.61 per 1,000 live births. The mortality risk of SIDS was significantly higher in infants with male gender and households classified by employment status as "other". Because the medical examiner system functioned only in the central urban area, most SIDS cases were diagnosed at autopsies in the urban area, whereas few cases were autopsied in the surrounding area. Compared to the urban area, the incidence of SIDS was significantly lower in the surrounding areas with low autopsy rates and with many death certificates of ambiguous death description.

Ambiguous description of the death scene and low autopsy rate may cause under-diagnosis of SIDS. Establishment of a nationwide medicolegal autopsy system and examination of the death scene is necessary to diagnose SIDS accurately.