Causes of nursery death beyond the neonatal period.

OBJECTIVE

To investigate causes of death in infants who died after 28 days, beyond the neonatal period but before discharge from the nursery, to establish their clinical courses and causes of death and to attempt to find criteria for earlier identification of these infants.

METHODS

We identified 30 such infants (12% of nursery deaths) from 1993 through 1998 and conducted a retrospective review of their records including placental pathology and autopsy reports when available. In all, 14 infants who weighed <or = 860 g at birth were matched with survivors.

RESULTS

The 30 infants divided almost equally into two groups. Of them 15 infants weighing >or = 880 g died of diverse congenital anomalies, including five with nonhemolytic hydrops and four with pulmonary hypoplasia. One infant without congenital anomaly weighed 3290 g. Support for this severely asphyxiated infant was withdrawn after 103 days because of progressive cortical atrophy. The remaining 14, the largest of which weighed 860 g, died of complications of prematurity, which we termed postponed neonatal deaths (PND). They followed a typical course of progressive multiple organ failure. All received assisted ventilation and postnatal steroids, developed chronic lung disease, and were on ventilation at the time of death. Renal insufficiency occurred late in the course. Acute infections and renal failure were the major proximal causes of death. When compared with surviving controls the PND were less likely to have received antenatal steroids and received more inotropic agents for cardiovascular support and more amphotericin for fungal infection; surgery for perforated bowel was confined to the PND.

CONCLUSIONS

The incidence of postneonatal nursery deaths has not changed in more than 20 years remaining at 11 to 12% of nursery deaths. Congenital anomaly was a prominent cause of death (50%). When infants without congenital anomalies (PND) were compared to surviving controls, no differences were found, which could reliably identify PND early in their course. The only potentially preventable factor was lack of antenatal steroid exposure in the PND.