OBJECTIVENosocomial infections (NIs) have become a matter of major concern in neonatal intensive care units (NICU). The objectives of this study were to investigate the incidence of nosocomial infections of newborn infants in NICU and to explore the risk factors and strategies of infection control.

METHODSThe study enrolled 638 hospitalized newborn infants from Apr 2003 to Dec 2004. The clinical data, such as the clinical manifestation, the condition of colonized bacteria, were collected and analyzed by using SPSS software.

RESULTThere were 88 times of nosocomial infections in 74 newborn infants. The overall incidence of nosocomial infections was 11.6%. The mean duration from admission to first episode of NI was 7.98 +/- 4.58 days. The incidence density was 14.9 per 1000 NICU patient-days. Catheter-correlated hematogenous infection rate was 18 per 1000 umbilical or central line-days; the ventilator-associated nosocomial pneumonia rate was 63.3 per 1000 ventilator days. The smaller the gestational age and the lower the birth weight, the higher the incidences of nosocomial infection. The duration of hospitalization was longer in these infected infants than those non-infected infants. Univariate analysis indicated that gestational age < or = 32 W, the parenteral nutrition, birth weight < or = 1500 g and mechanical ventilation, apnea, small for gestational age infant, central venous catheter (P < 0.05) were risk factors for NIs. Multivariate analysis identified 3 independent risk factions: the parenteral nutrition ([OR] = 7.185 [95% CI, 3.399 - 15.188]), birth weight < or = 1500 g ([OR] = 3.310 [95% CI, 1.100 - 9.963]) and mechanical ventilation ([OR] = 2.527 [95% CI, 1.092 - 5.850]). The most common infection was pneumonia (45.4%). The mortality rate of nosocomial infections was 4.1%. Bacterial surveillance was examined by nasopharyngeal and rectal swab culture immediately on hospital admission and then once a week. The incidence rate of NIs was 24.8% in patients whose nasopharyngeal and rectal swab culture indicated bacterial colonization, and 1.9% in patients without bacterial colonization (chi(2) = 79.7, P < 0.001).

CONCLUSIONIt is important to identify the high risk factors for nosocomial infections in newborn infants in NICU. Reducing the duration of the parenteral nutrition and the virulence manipulation as far as possible and getting the message of individual bacterial colonization in NICU may be conducive to decrease of the incidence of nosocomial infections and provide reference for rational clinical drug administration.