Does diagnosis influence end-of-life decisions in the neonatal intensive care unit?

To determine the influence of physiological status and diagnosis at the time of death on end-of-life care.

Retrospective descriptive study in a regional referral level IV neonatal intensive care unit (NICU) of infants who died from 1 January 1999 to 31 December 2008. Infants were categorized based on diagnosis (very preterm, congenital anomalies or other) and level of stability. Primary outcome was level of clinical service provided at end of life (care withheld, care withdrawn or full resuscitation).

From 1999 to 2008, there were 414 deaths in the NICU. Congenital anomaly was the leading diagnosis at the time of death, representing 45% of all deaths. Comparing mode of death, very preterm newborns were more likely than infants with congenital anomalies to have received cardio-pulmonary resuscitation (CPR) at the time of death (26% vs 13%, P < 0.01) and were significantly more unstable (75% vs 52%, P < 0.01). Infants aged 22 to 24 weeks were mostly unstable and significantly more likely to receive CPR than infants with any other diagnosis.

Over the 10-year period, very preterm infants were more likely to be physiologically unstable and to receive CPR at the time of death than infants with any other diagnosis. This finding was especially true for infants at the edge of viability (22 to 24 weeks). These differences in end-of-life care suggest that the quality of life and medical futility may be viewed differently for the least mature infants.