Multiple risk factors have an impact on the prevalence of CVC BSI in preterm neonates. We performed a prospective nation-wide study on the impact of process and structure parameters in the majority of German NICUs.
Our results demonstrate that high staffing levels are associated with a lower incidence of CVC BSI and laboratory-confirmed BSI (with organisms other than coagulase negative staphylococci). The results are congruent with several other studies on staffing [4, 11]. A study by Pittet et al. demonstrated that an increased workload is associated with diminished hand hygiene compliance . Cho et al. showed that even a small increase in the nurse-per-patient ratio is associated significantly decreased odds for adverse events . Other studies showed that understaffing as well as overcrowding is associated with a higher risk of outbreaks on NICUs [14, 15]. Two other studies did not observe an influence of nurse staffing [16, 17]. However, both studies assessed the situation on ICUs rather than NICUs. We did not assess the nurse per patient ratio, but we showed that compliance with the in-house recommendations on staffing levels has the potential to prevent healthcare-associated BSI and is a relevant quality assurance tool. So far, there is no general reference for staffing ratios on NICUs. However, the German Commission on Hospital Hygiene and Infection Prevention suggests at least high levels of appropriately trained nurses .
Birthweight and gestational age have been shown in the literature to be the predominant patient related risk factors for healthcare-associated infections [2, 19–21]. We could fully confirm these findings our study.
The preparation of infusions at laminar airflow benches has been recommended to minimize the risk of contamination [22, 23]. Thomas et al. demonstrated that the training background of the preparing person can be critical, rather than the preparation site . We did not assess the training background of personnel. Nevertheless, our data confirms the importance of laminar-flow benches in the preparation of intravenous fluids.
There are only few studies on the mode of delivery in VLBW newborns. It was found that cesarean section is associated with BSI in VLBW infants, while other studies reported improved morbidity and mortality [25–27]. Even though our results confirm the potential protection by cesarean section, this might also be an indirect effect of emergency vaginal delivery. Cesarean sections are usually handled in a well-prepared professional environment and more common in preterm infants .
This study has following limitations. It is based on the NEO-KISS database. Even though we thoroughly proofed their reliability, the accuracy of data is dependent on the quality of these data. We furthermore depend on the information we obtained by the inquiries that could be subject of relevant recall bias. We did not assess the nurse patient ratio and therefore cannot provide an exact threshold as benchmark for quality assurance. Nevertheless our results show that the planned numbers of staffing can have an impact on their rate of nosocomial infections. Planned staffing should therefore strictly be realized.