Objective To investigate the relationships between mortality of very low birth weight infants or very low gestational age infants and volume of activity, staff workload and geographical area in Italian neonatal intensive care units (NICU).

Methods NICU data were collected with a cross-sectional design with monthly repeated measures between May and December 2005. Data about the infants admitted from 1 January to 31 December 2005 were prospectively collected. The outcome of interest was mortality before discharge; its association with the NICU and subjects’ characteristics was studied by univariate and multivariate analysis.

Results The participating NICU were 105, and the babies enrolled were 4811. The mean values of gestational age and birth weight were 29.2 weeks (range 21–39) and 1194 g (range 300–2840). The NICU median volume of activity was 39 (range 4–133) and the daily infant to staff ratio ranged from 0 to 3.33. The overall mortality rate was 16.5%. When comparing the first with the fourth quartile of the volume and infant to staff ratio distributions, mortality decreased from 17.8% to 15.2% and from 17.6% to 14.7%, respectively. Mortality in the northern, central and southern regions was 12.2%, 18.6% and 20.5%. In the multivariate logistic analysis the explanatory variables associated with mortality were gestational age, birth weight, gender, geographical area and infant to staff ratio.

Conclusions In Italy perinatal and neonatal care are still largely ineffective, inefficient and characterised by profound geographical inequalities. The wide variation in the volume of activity suggests that a more appropriate and reliable national NICU certification system is required.