LENGTH OF STAY, COST AND MORTALITY OF HEALTHCARE-ACQUIRED BLOODSTREAM INFECTIONS IN CHILDREN AND NEONATES: A SYSTEMATIC REVIEW AND META-ANALYSIS

S. Karagiannidou, C. Triantafyllou, T. Zaoutis, V. Papaevangelou, G. Kourlaba  

37th Annual Meeting of the European Society for Paediatric Infectious Diseases (ESPID 2019)
Ljubljana, Slovenia, 6-11 May, 2019

BACKGROUND: Healthcare-associated infections (HAIs) are associated with increased mortality, length of stay (LoS) and healthcare cost. Healthcare-acquired bloodstream infections (HA-BSIs) are the most common HAI in children and neonates. The aim of this systematic review and meta-analysis was to present the attributable mortality, LoS and healthcare cost of HA-BSIs in children and neonates.

METHODS:  A systematic search up to September 2018 was conducted in PubMed, Cochrane, and CINAHL databases using combinations of a considerable number of relevant words). Cited references from selected articles were used to find additional studies that were not retrieved in the initial search. Studies eligible for inclusion were case-control or cohort studies published in English and available as full text that provided data for at least one of the following: attributable or excess mortality, healthcare cost, or LoS. Study quality was evaluated using the Critical Appraisal Skills Programme Tool (CASP). A meta-analysis was performed using fixed effects modelling.

RESULTS:  Of 4660 papers identified in the search, 21 were included in the systematic review and 13 in the meta-analysis. Attributable mortality was presented in 7 attributable healthcare cost in 9 and attributable LoS in 16 studies. It was found that the attributable mortality rate ranged between 1.43% to 24%, whereas the attributable healthcare cost ranged from $1642.16 to $160804 (2019 USD) per patient with HA-BSI. The attributable LoS ranged between 1.57 to 27.8 days. The wide range is due to differences in demographic characteristics among the study populations, as well as in the type of analysis performed (in the majority of studies matching regression was used), and in the study design (12 of them are prospective cohort and 9 are retrospective cohort studies). The results of the meta-analysis indicated that the pooled mean attributable hospital LoS was 16.91 days (95% CI= 13.70, 20.11). There was no heterogeneity between the studies (I2=27.7%), and no publication bias was detected (Egger’s bias test p=0.705). With regard to the overall attributable mortality, meta-analysis revealed that the pooled percentage was 0.08 (95% CI= 0.06, 0.09). There was not statistically significant heterogeneity between the studies (I2=45.3%, p=0.067). Moreover, meta-analysis showed that the pooled percentage of attributable Neonatal Intensive Care Unit (NICU) and Pediatric Intensive Care Unit (PICU) mortality was 0.08 (95% CI= 0.03, 0.13), and 0.13 (95% CI= 0.07, 0.19) respectively. What is more, the pooled adjusted mean attributable NICU LoS was 11.370 days (95% CI= 4.849, 17.891) and the pooled mean attributable PICU LoS was 16.399 days (95% CI= 10.087, 22.712). Last but not least, meta-analysis was not conducted for attributable cost due to the lack of eligible studies.

CONCLUSIONS: HA-BSIs in children and neonates are associated with higher mortality, LoS, and healthcare cost than is found among children and neonates without HA-BSI. This finding seems to justify and may enhance efforts to implement prevention strategies.