Abstract:
During an outbreak of Rotavirus (RV) on all neonatal units at the Department of Neonatology at the University Children’s Hospital in November 2002, RV was even detected in stool samples of asymptomatic neonates. This indicated a higher prevalence of RV infections on neonatal wards than previously assumed; we wanted to determine this prevalence. Furthermore, the hypothesis was tested that a higher sensitivity of RV detection can be achieved by establishing a Reverse Transcriptase – Polymerase Chain Reaction (RT-PCR) for RV. Besides, we examined if there are characteristic symptoms of infected neonates, how long neonates excrete the virus and if there is a link between RV infection and emergence of necrotizing enterocolitis (NEC).
Stool samples of all 123 patients admitted to the neonatal intensive care unit (NICU) between January 1, 2003 to April 30, 2003 were prospectively tested for RV. RV was detected in 28 neonates (23%) via Enzyme-linked immunosorbent assay (ELISA) and/or RT-PCR.
Virus shedding during the first days of life occurred in 6 newborns; RV was even detected in meconium. This indicates a possible perinatal transmission of the virus.
Agreement between the results of ELISA and RT-PCR was poor: only in 4 samples (out of 42) RV could be detected with both methods.
Characteristic symptoms for infection in neonates could not be identified. Neither a higher stool frequency nor a higher stool quantity, watery stools or signs of dehydration occurred more often in infected than in uninfected newborns.
There was no significant difference between term and preterm neonates. However premature infants seemed to have a higher likelihood of being infected with Rotavirus as well as a prolonged shedding of the virus. This can explain the high infection rates on the NICU and reveals the importance of taking precautions for prevention and infection control.
One of the premature neonates who became infected with RV subsequently developed a necrotising enterocolitis (NEC). It is not possible to draw conclusions out of this individual case but the possibility of a connection between RV infection and emergence of NEC should be considered.
Breastfeeding can be associated with a reduced risk of RV infection: The proportion of breast-fed neonates in the non infected control group was significantly higher than among the infected neonates.
The reason for the poor agreement between ELISA and RT-PCR results remains unclear. False positive results resulting from cross reactions in the ELISA with bacterial antigens and/or alterations of the viral RNA through the inevitable cycles of freezing and thawing have to be discussed, but cannot be proved on the basis of the data included in this study. Based of these results, neither the ELISA nor the RT-PCR can be considered as an optimal test for detection of RV infection in neonates.
Infections with RV are common in newborns. Despite the discrete symptoms, a relationship with other, more serious sequelae cannot be excluded. This and the high contagiosity argue in favor of a screening for RV infections. In contrast, however, the mild symptoms argue against a routine RV screening. Before a decision can be reached, further studies on detection methods for RV in neonates should be performed.